Вы находитесь на странице: 1из 354

The role of Math5 (Atoh7) in retinal and optic nerve development and

human disease
by
Lev Prasov

A dissertation submitted in partial fulfillment


of the requirements for the degree of
Doctor of Philosophy
(Human Genetics)
in the University of Michigan
2012

Doctoral Committee:
Associate Professor Thomas M. Glaser, Chair
Professor Sally A. Camper
Professor Pamela A. Raymond
Associate Professor Donna M. Martin
Associate Professor David L. Turner

Lev Prasov
2012

To my parents and grandparents,


for the genes and environment

ii

ACKNOWLEDGEMENTS
Many people have contributed to the work presented in this thesis. First
and foremost, I would like to thank my advisor, Tom Glaser, for his guidance and
support throughout this process. He deserves much of the credit for my success
as a graduate student, and has truly taught me to be a good and rigorous
scientist. I would like to thank my thesis committee (David Turner, Donna Martin,
Sally Camper, and Pamela Raymond) for fruitful discussions and constructive
feedback. I would like to thank all of the Glaser lab members past and present
(Chris Chou, Dellaney Rudolph, Mindy Nagy, Nate Vale, and Sherry Taylor) for
creating a vibrant and productive work environment, including the entertaining
discussions both about science and about nothing. I would also like to
acknowledge the efforts and support of all the people that have provided me with
reagents, shared ideas, assisted with experiments, and helped with manuscripts.
Foremost, I would like to thank Chris Chou, Joe Brzezinski, Dave Turner, and
Nadean Brown, whose advice and support have been truly invaluable. Id like to
thank the staff at the MSTP and HG department, especially Ron Koenig, Ellen
Elkin, and Karen Grahl, for their administrative support and advice. Id like to
thank my friends, who have helped me stay sane throughout the PhD process.
Finally, Id like to thank my parents and grandparents for their love and support,
and my brother, Zahar, for being my first and best teacher.

iii

PREFACE

The work presented in this thesis comprises most of my efforts in Glaser


laboratory since August of 2007. Many of the chapters have been accepted or
submitted for publication, and were done as a collaborative effort with other
scientists. Outside contributions and publication status are noted below.
Additionally, acknowledgements are appended to each chapter to recognize
other contributors.
Chapter II describes a critical analysis of Math5 (Atoh7) splicing in the
developing retina and brain. For this project, Nadean Brown conducted the
original Northern blot analysis and tested Math5 antibodies on retinal tissues.
This work was published as: Prasov L, Brown NL, Glaser T. A critical analysis
of Atoh7 (Math5) mRNA splicing in the developing mouse retina. PLoS One.
2010, 5(8), pii:e12315.
Chapter III describes an analysis of the lineage and cell cycle properties of
Math5 expression. Joe Brzezinski generated and characterized the Math5>Cre
BAC transgenic mice. He also collected the data for quantitative lineage analysis
in Math5 wild-type and knockout backgrounds, conducted the reporter
concordance, rod vs. cone birthdating experiments, and did some preliminary cell
cycle analysis. A modified version of this chapter has been accepted for

iv

publication as: Brzezinski JA IV, Prasov L, Glaser T. Math5 establishes an early


RGC competence state in a subpopulation of exiting retinal progenitor cells. Dev.
Biol. 2012.
Chapter IV describes cell cycle and clonal analysis that pinpoints the
timing of the RGC fate decision. Chapter V describes Math5 over-expression
experiments. These experiments were carried out by me, with technical support
in cloning and mouse colony management from Dellaney Rudolph and Sue Tarl.
Transgenic mice were generated by the UM Transgenic Animal Model Core, and
Mitchell Gillett assisted with histology of plastic sections. Modified versions are
being submitted for publication as Prasov L., Glaser T. Dynamic expression of
ganglion cell markers in retinal progenitors during the terminal cell cycle and
Prasov L, Glaser T. Pushing the envelope of retinal ganglion cell genesis:
context dependent function of Math5 (Atoh7).
Chapter VI describes the screening and identification of ATOH7 (MATH5)
mutations in congenital neural and vascular diseases of the retina. I designed
and carried out all of the functional analysis. Tehmina Masud and Edward Oliver
assisted with the identification of ATOH7 mutations, and screening of cases.
Aiysa Abid, Shagufta Khaliq, Qasim Mehdi, Eduardo Silva, Amy Lewanda, Mark
Borchert, Mehul Dattani provided DNA samples and clinical information for cases
screened in this study. Dan Kelberman and Jane Snowden screened ONH
cases for coding ATOH7 variants. Genome-wide Illumina-OmniQuad genotype
was done with Susan Dagenais and Bob Lyons at the UM DNA sequencing core.
A modified version of this chapter has been submitted for publication as Prasov L,

Masud T, Khaliq S, Mehdi SQ, Abid, AY, Oliver ER, Silva E, Brodsky MC,
Borchert M, Kelberman D, Snowden JC, Dattani M, Glaser T. Mutations in Atoh7
cause recessive persistent hyperplasia of the primary vitreous.

vi

TABLE OF CONTENTS
DEDICATION ....................................................................................................... ii
ACKNOWLEDGEMENTS ................................................................................... iii
PREFACE ........................................................................................................... iv
LIST OF FIGURES ............................................................................................... x
LIST OF TABLES ............................................................................................. xiv
ABSTRACT ........................................................................................................ xv
CHAPTER I: VERTEBRATE RETINAL DEVELOPMENT ................................. 1
Structure and composition of the retina ............................................................. 1
Control of retinal development........................................................................... 4
Cell divisions and the retinal cell cycle .............................................................. 6
Cell death in the retina ...................................................................................... 9
Intrinsic and extrinsic signals in retinal development ....................................... 10
Math5 (Atoh7) and control of RGC fate ........................................................... 13
Retinal vascular development and the role of RGCs ....................................... 15
Diseases of the optic nerve and retinal vasculature ........................................ 17
Novel insights into RGC development and the role of Math5 (ATOH7) ........... 20
CHAPTER II: A CRITICAL ANALYSIS OF MATH5 (ATOH7) mRNA
SPLICING IN THE DEVELOPING MOUSE RETINA ......................................... 29
Abstract ........................................................................................................... 29
Introduction...................................................................................................... 30
Materials and Methods .................................................................................... 32
Results ............................................................................................................ 40
Discussion ....................................................................................................... 50
Acknowledgements ......................................................................................... 55

vii

CHAPTER III: MATH5 DEFINES THE GANGLION CELL COMPETENCE


STATE IN A SUBPOPULATION OF RETINAL PROGENITOR CELLS
EXITING THE CELL CYCLE.............................................................................. 79
Abstract ........................................................................................................... 79
Introduction...................................................................................................... 80
Materials and Methods .................................................................................... 85
Results ............................................................................................................ 96
Discussion ..................................................................................................... 111
Acknowledgements ....................................................................................... 120

CHAPTER IV: DYNAMIC EXPRESSION OF GANGLION CELL MARKERS


IN RETINAL PROGENITORS DURING THE TERMINAL CELL CYCLE ........ 147
Abstract ......................................................................................................... 147
Introduction.................................................................................................... 148
Materials and Methods .................................................................................. 150
Results .......................................................................................................... 154
Discussion ..................................................................................................... 158
Acknowledgements ....................................................................................... 163

CHAPTER V: PUSHING THE ENVELOPE OF RETINAL GANLGION CELL


GENESIS: CONTEXT DEPENDENT FUNCTION OF MATH5 (ATOH7) ......... 173
Abstract ......................................................................................................... 173
Introduction.................................................................................................... 174
Materials and Methods .................................................................................. 178
Results .......................................................................................................... 189
Discussion ..................................................................................................... 203
Acknowledgements ....................................................................................... 211

CHAPTER VI: ATOH7 (MATH5) MUTATIONS CAUSE


AUTOSOMAL RECESSIVE PERSISTENT HYPERPLASIA OF
THE PRIMARY VITREOUS.............................................................................. 240
Abstract ......................................................................................................... 240
Introduction.................................................................................................... 241
viii

Patients and Methods.................................................................................... 245


Results .......................................................................................................... 254
Discussion ..................................................................................................... 262
Acknowledgements ....................................................................................... 268
CHAPTER VII: DISCUSSION AND FUTURE DIRECTIONS .......................... 286
The structure of Math5 (Atoh7)...................................................................... 286
The fate plasticity of Math5 (Atoh7) cells....................................................... 287
The non-autonomous role of Math5-expressing cells .................................... 290
The RGC fate decision and the role of Brn3b................................................ 292
Testing the pioneering model of RGC fate .................................................... 293
The role of ATOH7 (MATH5) in human disease ............................................ 294
Concluding remarks ...................................................................................... 295
REFERENCES ................................................................................................. 298

ix

LIST OF FIGURES

CHAPTER I:
Figure I-1. Structure of the eye and retina. ..................................................... 23
Figure I-2. Molecular mechanisms of retinal development ............................. 24
Figure I-3. Math5 (Atoh7) knockout mice have no optic nerves, and severe
deficiencies in RGCs ....................................................................................... 26
Figure I-4. Development of the retinal vasculature ......................................... 27
CHAPTER II:
Figure II-1. Anatomy of the Math5 transcription unit ....................................... 56
Figure II-2. Math5 messenger RNAs .............................................................. 58
Figure II-3. Math5 embryonic eye RT PCRs with increasing amounts of
betaine. ........................................................................................................... 60
Figure II-4. RT PCRs of Math5 RNA transcribed in vitro ................................ 61
Figure II-5. Triplex competitive RT PCR assay to evaluate trace levels of
Math5 splicing in the embryonic retina ........................................................... 63
Figure II-6. Ribonuclease protection assays................................................... 65
Figure II-7. Model explaining the observed results ......................................... 67
Figure II-S1. Math5 ESTs in the public domain .............................................. 68
Figure II-S2. Evaluation of Math5 antibodies. ................................................. 70
Figure II-S3. Math5 splicing in the cerebellum................................................ 72
Figure II-S4. Splicing patterns in the mouse Atonal related bHLH genes ....... 74
Figure II-S5. Secondary structure for Math5 mRNA ....................................... 75
CHAPTER III:
Figure III-1. Math5 is expressed by early retinal progenitors during
or shortly following their terminal cell cycle ................................................... 122
Figure III-2. Construction and expression of the Math5>Cre transgene ....... 124
Figure III-3. Math5+ progenitors contribute differentially to all retinal
cell types ....................................................................................................... 126
x

Figure III-4. All Math5>Cre progenitors express similar levels of Cre,


regardless of cell fate. ................................................................................... 128
Figure III-5. The fate distribution of Math5+ progenitors changes over time. 130
Figure III-6. Math5 marks many of the earliest born cells in the retina ......... 132
Figure III-7. A subset of Brn3b+ RGCs derives from the Math5 lineage ....... 134
Figure III-8. Retrovirally marked clones exhibit symmetric and asymmetric
patterns of Math5 expression ........................................................................ 135
Figure III-9. Natural history of the Math5 lineage.......................................... 137
Figure III-S1. Copy number and integrity in Math5>Cre transgenes............. 140
Figure III-S2. Birthdating curves for rods and cones .................................... 141
Figure III-S3. Proneural bHLH factors Neurod1 and Math5 are expressed
in overlapping subsets of progenitor cells during early retinal neurogenesis. 142

CHAPTER IV:
Figure IV-1. Timing of cell cycle progression in the mouse retinal
neuroepithelium at E13.5 and E15.5 ............................................................. 165
Figure IV-2. Coexpression and onset analysis of amacrine and horizontal
markers Ptf1a and AP2. .............................................................................. 166
Figure IV-3. The onset of Brn3b and Isl1 expression within individual
cells is progressively delayed during retinal development. ............................ 168
Figure IV-4. Co-expression of Brn3b and Isl1 during or shortly after the
terminal cell cycle .......................................................................................... 170
Figure IV-5. Paired ganglion cells can be generated from retinal
progenitors by symmetric terminal division.................................................... 168
CHAPTER V:
Figure V-1. Crx and Math5 (gal) are expressed in overlapping subsets of
cells shortly after cell cycle exit ..................................................................... 213
Figure V-2. Characterization of the Crx>Math5-IRES-Cre and Crx>Cre
conventional and BAC transgenic mice ......................................................... 214
Figure V-3. Widespread Crx>Math5 expression has little effect on cell fate
decisions in the retina.................................................................................... 216
Figure V-4. Widespread Crx>Math5 expression does not alter RGC
abundance or retinal histology....................................................................... 217
Figure V-5. Widespread Crx>Math5 expression does not extend the profile
of RGC births, but decreases the numbers of early-born photoreceptors ..... 218
Figure V-6. Retroviral Math5 overexpression does not stimulate RGC fate
or cell cycle exit in retinal explant cultures .................................................... 219
xi

Figure V-7. Crx>Math5 expression partially rescues RGC fate


specification and optic nerve development in Math5 knockout (KO) mice..... 221
Figure V-8. RGC birthdates in transgenic, Math5 KO and rescued animals . 223
Figure V-9. Survival and generation of late-born RGCs are inhibited in
rescued animals ............................................................................................ 225
Figure V-S1. Consistent retinal expression patterns for multiple,
independent Crx>Cre Tg and Crx>Cre BAC transgene insertions ................ 227
Figure V-S2. Crx transgenic and BAC expression patterns in the
pineal gland, retinal pigmented epithelium, and ciliary body ......................... 228
Figure V-S3. Direct comparison of Crx>Math5 BAC and endogenous
Crx transcript levels ....................................................................................... 229
Figure V-S4. Math5 does not grossly alter secondary fate choices of
photoreceptors and bipolar cells.................................................................... 230
Figure V-S5. Cell fate spectrum of the conventional Crx>Math5 transgene . 231
Figure V-S6. Dual-reporter concordance experiment ................................... 233
Figure V-S7. The Crx>Math5 BAC transgene is expressed at low levels in
proliferative retinal progenitors ...................................................................... 234
Figure V-S8. Reduced photoreceptor births at E12.5 and E13.5 in
Crx>Math5 Tg mice ....................................................................................... 235
Figure V-S9. Crx>Math5 Tg mice and control mice exhibit similar levels of
apoptosis throughout development ............................................................... 236
Figure V-S10. RGC axons in Math5 KO and transgene-rescued mice
exhibit severe pathfinding defects ................................................................. 237
Figure V-S11. Brn3b expression in Crx>Math5 mice.................................... 238
CHAPTER VI:
Figure VI-1. The ATOH7 p.Asn46>His allele segregates with autosomal
recessive persistent hyperplastic primary vitreous (arPHPV) disease........... 270
Figure VI-2. The ATOH7 p.Arg65>Gly allele in a child with optic nerve
aplasia and developmental delay. ................................................................. 271
Figure VI-3. Sequence alignment and structural modeling of ATOH7
mutations ....................................................................................................... 272
Figure VI-4. The arPHPV mutant ATOH7 polypeptide (N46H) does
not bind DNA or activate transcription, while A47T and R65G variants
retain these functions .................................................................................... 274
Figure VI-5. Human ATOH7 R65G and A47T variants rescue ganglion
cell specification in Atoh7 -/- retinal explants, but N46H and L56P
mutants do not............................................................................................... 276
xii

Figure VI-S1. ONA Patient 1 carries a duplication that disrupts CNTN4 ...... 278
Figure VI-S2. The chromosome 3p26 duplication in Patient 1 is capable of
producing a truncated CNTN4 mRNA ........................................................... 279
Figure VI-S3. Chromosome 14q23 deletion in optic nerve aplasia
Patient 2 encompasses the OTX2 gene ........................................................ 281
Figure VI-S4. ATOH7 variants have similar protein stability ......................... 282
Figure VI-S5. Low power views of retinal explant rescue experiments ......... 283

CHAPTER VII:
Figure VII-1. Outline of the Mosaic Analysis of Double Markers (MADM)
strategy ......................................................................................................... 297

xiii

LIST OF TABLES

CHAPTER II:
Table II-S1. Oligonucleotide primers in this study .......................................... 76
Table II-S2. PCR conditions in this study ....................................................... 77
Table II-S3. DNA sequence flanking deletions in RT-PCR products .............. 78
CHAPTER III:
Table III-1. Cell type distribution of Math5 lineage descendants in
wild-type Math5>Cre transgenic retinas ........................................................ 139
Table III-S1. Cell type distribution of Math5 lineage descendants in Math5
mutant mice ................................................................................................... 143
Table III-S2. Dual reporter concordance for Math5>Cre labeled
retinal cells .................................................................................................... 144
Table III-S3. Cumulative BrdU labeling experiment (E10.5 to P0) ................ 145
Table III-S4. Birthdates of Math5 lineage retinal descendants ..................... 146
CHAPTER V:
Table V-S1. Oligonucleotide primers and PCR conditions used
in this study. ................................................................................................... 239
CHAPTER VI:
Table VI-S1. Clinical Features of optic nerve aplasia cases......................... 284
Table VI-S2. Oligonucleotide primers and PCR conditions used
in this study ................................................................................................... 285

xiv

ABSTRACT
Vertebrate retinal histogenesis is controlled by both intrinsic transcriptional
programs and the microenvironment. The basic helix-loop-helix (bHLH) factor
Math5 (Atoh7) is required for differentiation of retinal ganglion cells (RGC), which
form the optic nerve. Math5 knockout mice lack RGCs, but only 10% of Math5expressing progenitors adopt the RGC fate, and only 55% of RGCs are lineal
descendents of Math5+ cells.
To define the role of Math5 in RGC development, I characterized the
transcriptional anatomy of mouse Math5, and showed that it is an unspliced,
single-exon gene, contrary to a recent high-profile report. I then tested the
contribution of Math5-expressing cells to the earliest born cohort of mouse retinal
neurons, which consist primarily of RGCs (~80%). Unexpectedly, I found that
only 20-30% of this cohort expresses Math5, yet most early RGCs depend on
Math5 function, suggesting a non-autonomous role for Math5-expressing cells in
RGC specification.
Next, I evaluated the onset of Math5 expression, and that of RGC markers
Brn3b and Isl1, with respect to the terminal cell cycle. Surprisingly, these
markers were expressed by neurogenic cells prior to terminal mitosis during early
development (<E14), but restricted to post-mitotic cells during later stages. By
retroviral clone analysis, I confirmed that early neurogenic cells often divide

xv

symmetrically, leading to paired RGC daughters. Retinal fate determination is


thus not strictly synchronized to cell cycle exit.
I then evaluated whether Math5 can bias terminally mitotic progenitors
toward the RGC fate. I broadly over-expressed Math5 using BAC and conventional transgenes controlled by Crx regulatory DNA. Unexpectedly, I found that
ectopic Math5 did not alter cell fate in a wild-type environment, but partially
rescued RGC development in Math5 mutant retinas. Early (pioneering) RGCs
are deficient in these mice, and rescue was incomplete. Transgene-derived lateborn RGCs exhibited pathfinding defects and were prone to apoptosis.
Finally, I evaluated the role of ATOH7 (MATH5) in human optic nerve
aplasia and hypoplasia, and familial persistent hyperplastic primary vitreous
(PHPV) disease. I identified a basic domain mutation (p.N46>H) in PHPV, and
established causation, using biochemical and functional assays. Together, these
studies provide important insights into the function of Math5 and RGC
development.

xvi

CHAPTER I
VERTEBRATE RETINAL DEVELOPMENT

Structure and composition of the retina


The vertebrate eye is a complex tissue that transmits visual information
from the environment to the central nervous system. Light enters the eye
through the cornea, is focused by the lens, and strikes the neural retina (Fig. I1A). The retina converts an electromagnetic signal (light) into an electrical and
neurochemical signal that can be processed by the central nervous system
(CNS) (Rodieck, 1998). The neural retina comprises six major classes of
neurons and one type of glia (Fig. I-1B). The neurons include rod and cone
photoreceptors; horizontal, bipolar, and amacrine interneurons; and retinal
ganglion cells. These form an organized structure with three layers of cell
bodies, including the outer nuclear layer (ONL), the inner nuclear layer (INL), and
the ganglion cell layer (GCL). The cellular layers are separated by two fiber
layers called the outer plexiform layer (OPL) and inner plexiform layer (IPL),
respectively, which contain precisely stratified synaptic connections between
photoreceptor endfeet or interneuron terminals and dendrites.
The ONL contains the cell bodies of ciliary photoreceptors, the major class
of cells that perceive light. They are subdivided into rods and cones, based on

cellular morphology, physiology, and role in the visual system. In most


mammals, particularly nocturnal species, rods are the most abundant cell class in
the retina and have relatively uniform properties (Jeon et al., 1998). These
photoreceptors sense low-levels of light. In contrast, cones are responsible for
color vision, and have heterogeneous spectral properties. Cones sense
characteristic wavelengths of light, depending on the opsin isotype that they
express. In primates, cones form a high-density cluster in the central retina
called the fovea or macula, an avascular region that provides the highest acuity
color vision (Yamada, 1969; Yuodelis and Hendrickson, 1986). In response to
light, a cyclic GMP cascade is triggered by photoisomerization of a retinoid
chromophore and conformational change in the photopigment protein (rhodopsin
in rods, and L-, M-, or S-opsin in cones of trichromatic animals), leading to
graded membrane hyperpolarization and stimulation of downstream bipolar
neurons (Yau and Baylor, 1989).
Bipolar cell interneurons reside in the outer aspect of the INL. Their
primary function is to transmit electrical signals from photoreceptors to ganglion
cells, which reside in the GCL (Rodieck, 1998). Amacrine and horizontal
interneurons also occupy characteristic positions in the inner and outer INL,
respectively, and have distinct roles in the processing of visual information.
These cells modulate the electrical signals of other neurons through excitatory or
inhibitory inputs, enhancing spatial integration and contrast detection (Rodieck,
1998). Mller glia cells also have somata that are located in the INL, but their
processes extend radially across the entire retinal thickness, from the outer

limiting membrane (OLM) to the inner limiting membrane (ILM) (Fig. I-1B). These
glial cells are important for maintaining the physiology and laminar structure of
the retina (Bringmann et al., 2006; Willbold et al., 2000).
Bipolar cells synapse on retinal ganglion cells (RGCs), whose cell bodies
reside in the GCL. RGC axons coalesce to form the optic nerves, which project
to areas in the central nervous system (CNS) that are responsible for visual
processing, and autonomic tasks, including eye positioning and object tracking,
modulating sensitivity to different levels of light, and controlling circadian rhythm
(Rodieck, 1998). These areas include the superior colliculi (SC), the lateral
geniculate nuclei (LGN), intergeniculate leaflets, the pretectum, the accessory
optic system (AOS), and the suprachiasmatic nuclei (SCN) (Schiller and Malpeli,
1977; Simpson, 1984). A small number of RGCs have intrinsic photosensitivity
(ipRGCs), express the photopigment melanopsin, project to the SCN, and are
important for photoentrainment of circadian rhythms (Gooley et al., 2001; Hattar
et al., 2002; Provencio et al., 1998). In addition to RGCs, the mammalian GCL
contains displaced amacrine cells (Masland, 1988), which have similar functions
as their INL counterparts.
There is significant diversity in the retina beyond the seven major cell
types (Masland, 2001), with particularly many subclasses of amacrine (MacNeil
and Masland, 1998), bipolar (Ghosh et al., 2004; Kim et al., 2008) and ganglion
cell (Rockhill et al., 2002) neurons. The relative abundance of each retinal cell
type is also quite different (Jeon et al., 1998). In particular, the ratio between
photoreceptors (input) and RGCs (output) highlights the importance of signal

integration, and varies widely across the retina, giving different levels of spatial
resolution. The overall number of photoreceptors vastly exceeds the number of
bipolar cells (~11:1), which in turn exceeds the number of ganglion cells (~12:1).
This integration greatly increases the sensitivity of the retina.

Control of retinal development


The vertebrate retina develops in two major phases. During the
morphogenetic phase, the optic vesicles evaginate from the ventral diencephalon
and invaginate to form bilayered optic cups (Chow and Lang, 2001; Spemann,
1901). Initially, there is an open cleft in the ventral retina, termed the optic or
choroid fissure. Gradually, this closes during early development through a
combination of cell signaling and adhesion (Barishak, 1992; Gregory-Evans et
al., 2004). In the histogenic phase, beginning at embryonic day E11 in the
mouse or 5th week in human gestation, some retinal progenitors (RPCs) begin to
exit the cell cycle and differentiate, while most continue to proliferate to expand
the progenitor pool.
Pioneering birthdating studies, in which [3H]-thymidine was used to identify
terminal progenitors by labeling the last phase of DNA replication, have
established an invariant, but overlapping, histogenic birth order across all
vertebrate species (Carter-Dawson and LaVail, 1979; Rapaport et al., 2004;
Sidman, 1961; Young, 1985a) (Fig. I-2A). At the onset of neurogenesis, ganglion
cells are the first cell type to exit the cell cycle and differentiate. In the mouse,
RGC birthdates initiate at E11, peak at E14, and terminate by P0 (Drager, 1985;

Young, 1985a). This temporal profile overlaps significantly with that of cone,
horizontal, amacrine cells. Rods, Mller glia and bipolar cells have
characteristically later birthdates, which peak in the perinatal or neonatal periods
in rodents.
Clonal analyses of retinal progenitors in frog and rodents, using retroviral
or plasmid vectors with histochemical markers, have revealed that the seven
major retinal cell types are generated from a common progenitor pool (Holt et al.,
1988; Turner and Cepko, 1987; Turner et al., 1990; Wetts and Fraser, 1988). In
these studies, composition and size among various clones was largely
heterogeneous, suggesting that no strictly determined lineages exist in the retina,
unlike those observed in the C. elegans nervous system (Ruvkun, 1997).
Indeed, large clones could contain both early and late cell types. Additionally,
the existence of two-cell clones with discordant fates suggests that commitment
to a particular cell fate must occur during or after the terminal division.
Heterochronic co-culture and transplantation experiments in rodents
likewise provided clues towards the mechanism of retinal fate determination.
When late (post-natal) progenitors were mixed at low density with a large
number of early cells (early embryonic), these late progenitors were incapable
of adopting early fates, suggesting an intrinsic restriction in competence, i.e. the
ability to respond to environmental signals (Belliveau and Cepko, 1999; Reh,
1992; Watanabe and Raff, 1990). In reciprocal experiments, early progenitors
were capable of adopting late fates, such as rod photoreceptor fate (Reh, 1992;
Watanabe and Raff, 1990). However, these RPCs were still heavily biased

towards early fates. In co-culture and transplantation experiments in frogs,


early progenitors were incapable of adopting late fates, suggesting that
competence in this species is temporally restricted (Rapaport et al., 2001).
The observations from heterochronic co-culture, clonal analysis, and
birthdating experiments have led to the development of a temporal competence
model for retinal development (Cepko et al., 1996; Livesey and Cepko, 2001;
Reh and Cagan, 1994; Wong and Rapaport, 2009) (Fig. I-2B). In this model,
progenitors progress through distinct intrinsic competence states in which RPCs
can adopt one or a small number of cell fates. During development, RPCs either
exit the cell cycle or continue to divide and progress to a new competence state.
The decision to stop dividing, and the ultimate histotypic fate, for cells in a
particular competence state is largely determined by environmental (extrinsic)
signals.
Alternatively, this set of observations is also consistent with a progressive
restriction model, similar to that proposed for the nervous system (Desai and
McConnell, 2000). In this model, early progenitors are multipotent, but heavily
biased toward the selection of early fates. As development proceeds, this bias is
altered by the changing environment, and RPCs are gradually restricted in their
selection of fates as the diversity of histotypic options narrows.
Cell divisions and the retinal cell cycle
In parallel with large changes in cell fate trajectory, there are alterations in
temporal parameters of the progenitor cell cycle, and the mode of cell division.
During early development, neural progenitors rapidly expand by employing a

symmetric self-renewing mode of division (P-P), in which both daughters


continue to proliferate (Gotz and Huttner, 2005; Huttner and Kosodo, 2005; Lu et
al., 2000). At the onset of neurogenesis, some progenitors give rise to daughter
cells that permanently exit the cell cycle and differentiate into neurons. The vast
majority of these divisions are asymmetric or stem-like (P-N), such that one
daughter continues to proliferate, while the other differentiates. In contrast, very
few symmetric terminal divisions occur (N-N), as these counteract the expansion
of the progenitor pool. As development proceeds, the balance of divisions
progressively shifts towards symmetric neurogenic divisions, which give rise to
two post-mitotic daughter cells (N-N). It is unclear what role, if any, the mode of
division (N-N vs. P-N) has on the final fate choice.
The kinetic parameters of the progenitor cell cycle also change
significantly over developmental time. The eukaryotic cell cycle is anchored by
two major events: S phase, in which DNA is replicated, and M phase, in which
mitosis and cytokinesis occur (Nurse, 2000). These are separated by two gap
phases, G2 and G1, respectively. Generally, G1 and S phases are the longest in
the progenitor cell cycle, while cells rapidly progress through G2 and M phases
(Young, 1985b). In the developing vertebrate retina, elongated progenitors span
the pseudostratified neuroepithlium. Their nuclei migrate along radial spindles in
a stereotypic oscillation, as cells progress through the cell cycle, in a process
known as interkinetic nuclear migration (Baye and Link, 2008) (Fig. I-2C). Nuclei
typically reside at the base of the retina (vitread surface) during DNA replication,
quickly migrate towards the apical surface during G2, undergo mitosis at the

apical (sclerad) surface of the retina, and return to the base in G1. After terminal
M phase, differentiating neurons migrate to their final laminar positions in the
neuroepithelium.
The dynamics of the RPC cell cycle have been extensively characterized
in rodents and other vertebrate species by window and cumulative nucleoside
labeling, cell counting, and percent labeled mitosis (PLM) methods (Alexiades
and Cepko, 1996; Fujita, 1962; Li et al., 2000; Sinitsina, 1971; Young, 1985b).
The common conclusion from these studies is that the overall cell cycle
progressively lengthens during development. This increase is largely attributed
to prolonged G1 and S phases of the cell cycle (Alexiades and Cepko, 1996;
Young, 1985b). Though much is known about the kinetics of the RPC cell cycle,
it remains unclear what role cycle dynamics play in the ultimate cell fate choice.
It is clear, however, that the cell cycle exit and fate determination are
correlated. Genetic overexpression or loss of cell cycle regulators, such as p27,
p57, Rb, cyclinD1, can have wide-spread effects on the retina, including the
distribution of retinal cell types (Das et al., 2009; Dyer and Cepko, 2000; Fantl et
al., 1995; Ohnuma et al., 1999; Zhang et al., 2004). Mitogenic factors, such as
Notch or Shh signaling molecules, tend to promote generation of late fates
(Jadhav et al., 2006; Levine et al., 1997); while loss of these and the presence of
neurogenic factors promote the generation of early fates (Dyer et al., 2003;
Riesenberg et al., 2009; Yaron et al., 2006). Despite these large influences, cell
cycle progression is not required for the generation of a wide array of neuronal
cell types in frog retina (Harris and Hartenstein, 1991). Thus, it remains to be

determined whether many of these effects are secondary to differences in the


retinal microenvironment or to the reduced number of RPCs during late stages of
development.
Despite many studies on fate determination and the cell cycle, the precise
timing of the fate decision relative to cycle exit is unknown. Some cells, such as
VC1.1+ amacrine precursors, may lose responsiveness to environmental signals
prior to terminal M phase (Belliveau and Cepko, 1999), similar to progenitors in
the ferret cortex (McConnell and Kaznowski, 1991). However, post-mitotic
precursors can also be biased toward other fates by intrinsic or environmental
signals (Adler and Hatlee, 1989; Brzezinski et al., 2010; Oh et al., 2007). Given
this heterogeneity in environmental response, it remains unclear whether the fate
decisions are made before or after terminal mitosis, and whether this property
varies among different cell types and different developmental stages.

Cell death in the retina


Programmed cell death is an additional regulatory mechanism that
ensures proper ratios of each cell type in the retina. During development, some
cell types are produced in vast excess. Half of the neurons in the GCL and INL
are culled during development (Voyvodic et al., 1995; Young, 1984). Inhibition of
programmed cell death pathways, by overexpression of Bcl-1 or deletion of Bax,
can significantly reduce this culling, indicating that canonical apoptosis is the
major regulatory mechanism (Martinou et al., 1994; Pequignot et al., 2003). The
death of the ganglion cells has been most extensively examined. It is estimated

that 50-70% of RGCs undergo cell death, with peak apoptosis in the neonatal
period (Crespo et al., 1985; Erkman et al., 2000; Farah and Easter, 2005; GalliResta and Ensini, 1996; Strom and Williams, 1998; Young, 1984). Activitydependent axon competition and defects in pathfinding are thought to be the
major mechanism underlying this neonatal cell death (Fawcett et al., 1984;
O'Leary et al., 1986; Scheetz et al., 1995).

Intrinsic and extrinsic signals in retinal development


The relative contribution of an intrinsic transcription program (clock) and
extrinsic signals in fate specification is heavily debated. Dissociated rat retinal
progenitors can produce a similar composition of progeny as those in the intact
retina, suggesting that progenitors retain at least a rudimentary intrinsic program
(Cayouette et al., 2003). This program is largely stochastic, but can generate the
appropriate numbers of various cell types (Gomes et al., 2011). Similarly,
progenitors in the frog retina appear to follow a rigidly fixed order of cell genesis
within clones, such that a single progenitor is only competent to produce only one
cell type at a given time (Wong and Rapaport, 2009). These results suggest that
a hard-wired intrinsic program is the dominant mechanism for retinal
development.
Despite the prominent role of intrinsic factors in specification, signals from
the retinal environment have a tremendous impact on the ultimate fate choice,
and can effectively override the intrinsic program. The retina has adapted many
negative feedback signals to ensure proper ratios of cell types. For example,

10

Delta ligands and secreted Shh, which are produced by nascent RGCs, are
known to negatively RGC genesis (Austin et al., 1995; Belliveau and Cepko,
1999; Waid and McLoon, 1998; Wang et al., 2005; Zhang and Yang, 2001).
These same signaling pathways (Notch and Hedgehog) also promote the
generation of other cell types (Furukawa et al., 2000; Jadhav et al., 2006; Levine
et al., 1997; Shkumatava et al., 2004). It is unclear, however, whether these
effects on cell fate are secondary to altering general progenitor dynamics (Austin
et al., 1995; Jensen and Wallace, 1997; Levine et al., 1997). For example, the
mitogenic and fate effects of Notch signaling are intertwined. Inhibition of the
Notch pathway may promote early fates simply by causing premature cell cycle
exit, increasing the fraction of cells that are specified in an early environment
(Austin et al., 1995; Nelson et al., 2007). Additionally, extrinsic factors can have
dual roles. They can support the differentiation of a particular cell type, while
inhibiting specification of new cells. For example, Shh produced in RGCs can
also promote axon pathfinding among nascent ganglion cells (Kolpak et al.,
2005; Sanchez-Camacho and Bovolenta, 2008). A vast array of other signaling
molecules, including ciliary neurotrophic factor (CNTF), fibroblast growth factor
(FGF), and transforming growth factor (TGF), also actively fine tune the
production of different cell types in the retina (Cepko, 1999; Ezzeddine et al.,
1997; Kim et al., 2005; Yang, 2004).
In an integrated view of retinal histogenesis, cell fate is fundamentally
controlled by intrinsic transcriptional cascades, which are influenced by the
environment (Fig. I-2D). A critical point in a progenitors life history is the

11

decision to exit the cell cycle. Thus, several core homeodomain (HD) and basichelix-loop-helix (bHLH) transcription factors, such as Chx10, Pax6, and Hes1,
are thought to control the progenitor proliferation and maintain RPCs in an
undifferentiated state (Ashery-Padan and Gruss, 2001; Marquardt, 2003;
Takatsuka et al., 2004).
Histotypic differentiation, the acquisition of unique cellular morphology and
functional features, is controlled by downstream transcriptional hierarchies, which
are similarly complex (Fig. I-2D). Although precise control of specification
remains largely unclear, many insights have emerged in recent years regarding
the hierarchy of differentiation from analysis of gene expression in loss- and gainof-function animal models. For example, in the development of inhibitory
amacrine and horizontal cells, it clear that both the HD factor FoxN4 and bHLH
factor Ptf1a (P48) are necessary for the development of these cell types, and
also bias progenitors toward this fate (Fujitani et al., 2006; Li et al., 2004). The
pattern of molecular epistasis (Lehner, 2011) in reciprocal mutants suggests that
FoxN4 precedes Ptf1a in the pathway. Loss of Ptf1a does not affect FoxN4
expression, but loss of FoxN4 impairs Ptf1a expression. Thus, a clear hierarchy
emerges, in which FoxN4 functions upstream of Ptf1a (Fujitani et al., 2006).
Similar transcriptional hierarchies have been established for the differentiation of
photoreceptors and other cell types (Ohsawa and Kageyama, 2008; Swaroop et
al., 2010). However, basic mechanisms controlling the fate choice, i.e. the
events that trigger these cascades, remain largely unknown.

12

Math5 (Atoh7) and control of RGC fate


The identification of the fly atonal homolog Math5 (mammalian atonal
homolog 5), also known as Atoh7 (atonal homolog 7) in vertebrate species
provided a critical insight into the mechanism of RGC development. In
Drosophila, atonal is vital for the specification of the first retinal neuron in the fly
ommatidium, the R8 photoreceptor (Jarman et al., 1994), which allows formation
of photoreceptors R1-R7 through a series of cell-cell inductive interactions. In
the vertebrate eye, this function of atonal is conserved, in that RGCs are the firstborn cell type. Homologs of atonal have been identified in the frog (Kanekar et
al., 1997), mouse (Brown et al., 1998), chicken (Liu et al., 2001; Matter-Sadzinski
et al., 2001), and human genomes (Brown et al., 2002). In mice, Math5 is
transiently expressed in a small population of retinal precursors in a pattern that
closely follows the RGC birthdating curve, with onset at E11 and cessation by
birth (Brown et al., 1998; Brown et al., 2001).
Atoh7 (Math5) is necessary for the development of more than 95% of
ganglion cells (Brown et al., 2001; Wang et al., 2001) (Fig. I-3). Although RGCs
represent on 0.5% of adult retinal cells (Jeon et al., 1998), Atoh7 mutant mice
have significantly thinner retinas (Brzezinski et al., 2005), ostensibly due to loss
of mitogenic effects from RGC-derived Shh (Dakubo and Wallace, 2004; Jensen
and Wallace, 1997; Levine et al., 1997). In the absence of RGC axons, there are
essentially no optic nerves. Because the optic nerves convey all photic stimuli to
the brain, mutant mice have no pupillary light reflex and their endogenous
circadian rhythms do not entrain to light (Brzezinski et al., 2005; Van Gelder et

13

al., 2003; Wee et al., 2002). Because the inner retina is thin, and bipolar neurons
are deficient, the mutants have abnormal retinal electrophysiology, with reduced
b-wave amplitudes in electroretinogram (ERG) recordings (Brzezinski et al.,
2005). Math5 -/- mice retain each of the other 7 major cell types, indicating that
Math5 is only required for ganglion cell development. Similarly, in zebrafish, the
ortholog ath5 is required for RGC development, as a point mutation in this gene
(lakritz) causes agenesis of RGCs and optic nerves (Kay et al., 2001).
Although Math5 is required for RGC development, it acts as a permissive
factor for RGC genesis. Lineage-tracing studies using a Math5-Cre knock-in
allele and BAC transgene have shown that Math5 descendents comprise most
major cell types in addition to RGCs (Brzezinski and Glaser, 2004; Feng et al.,
2010; Yang et al., 2003). Likewise, Math5 over-expression studies in chick and
frog suggest that this factor is not instructive for RGC fate. In both species,
overexpression of orthologs Xath5 and Cath5 in proliferating RPCs during early
development mildly biases progenitors toward the RGC fate (Brown et al., 1998;
Kanekar et al., 1997; Liu et al., 2001). However, when Xath5 is misexpressed
during late developmental stages in frogs, it promotes other neuronal fates
(Moore et al., 2002). In general, high-level expression of proneural bHLH factors
may drive cell cycle exit and neuronal differentiation (Farah et al., 2000). Thus,
the observed shifts in fate distribution may largely be due to this property of
bHLH factors, rather than a specific role of ath5.
The hierarchy of RGC differentiation remains incompletely understood
despite many genetic and genomic studies (Mu and Klein, 2004). It is clear that

14

POU-domain Brn3b (Pou4f2) and LIM-domain Isl1 (Islet-1) transcription factors


form two important regulatory nodes in ganglion cell specification and
differentiation (Mu et al., 2008; Pan et al., 2008). These factors, which are
thought to function downstream of Math5, are necessary for RGC development
and have distinct, but overlapping, transcriptional targets (Erkman et al., 1996;
Gan et al., 1996; Mu et al., 2008; Pan et al., 2008; Xiang, 1998). Brn3b is likely
to act as an instructive factor for RGC specification. It is expressed exclusively in
most developing RGCs and biases progenitors toward the RGC fate when
ectopically expressed (Badea et al., 2009; Feng et al., 2011; Liu et al., 2000; Qiu
et al., 2008). Two closely related paralogs, Brn3a and Brn3c, are functionally
interchangeable with Brn3b at the protein level (Liu et al., 2000; Pan et al., 2005).
However, their distinct spatiotemporal expression patterns confer unique roles in
RGC differentiation (Badea et al., 2009; Wang et al., 2002a; Xiang et al., 1995;
Xiang et al., 1993). Transcriptional profiling of Brn3b and Math5 knockout retinas
has led to the identification of other factors that function downstream in RGC
differentiation, such as HD protein eomesodermin (Eomes/Tbr2) (Mao et al.,
2008a; Mu et al., 2004). Despite these advances, much remains to be learned
about the factors that control the RGC fate decision and downstream
differentiation events.
Retinal vascular development and the role of RGCs
The retina is supplied by two major vascular systems (Fruttiger, 2007;
Gariano and Gardner, 2005; Provis, 2001). The outer retina is supplied by
diffusion from high-flow, fenestrated vessels in the choroid layer. The inner

15

retina is supplied by the hyaloid artery, which enters the eye through the optic
stalk and the embryonic fissure during early development (Zhu et al., 1999). The
choroidal and hyaloid vessels branch from the dorsal ophthalmic artery. The
hyaloid artery extends into the vitreous and forms the hyaloid vascular network,
which nourishes the lens and the inner retina (Fig. I-4). In order to accommodate
vision through an optically clear media, this vasculature remodels during the
early post-natal period in mice or late gestation in humans (Fruttiger, 2007;
Gariano and Gardner, 2005). The hyaloid vasculature regresses by apoptosis of
vascular endothelial cells, which is triggered by macrophage phagocytosis (Lang
et al., 1994). In turn, the blood vessels sprout by angiogenesis, following a
migrating astrocyte network that initiates at the optic stalk and spreads radially
along the inner surface of the retina (Fruttiger, 2007). These surface vessels
then dive into the retina to form two deep vascular plexi that surround the INL
and supply the inner retina. Astrocytes, marked by platelet-derived growth factor
receptor alpha (PDGFR) expression, form a scaffold for angiogenesis and are
important for proper development of the vasculature (Dorrell et al., 2002; Laterra
et al., 1990; Watanabe and Raff, 1988). RGCs elaborate platelet-derived growth
factor A ligand (PDGFRA), sonic hedgehog (Shh), and other trophic molecules
that are critical for astrocyte proliferation and migration (Dakubo et al., 2003;
Fruttiger et al., 1996; Fruttiger et al., 2000).
The retinal vasculature develops abnormally in Atoh7 mutant mice lacking
RGCs (Brzezinski et al., 2003; Edwards et al., 2011) (Fig. I-4). In these mice, the
mature retinal vasculature fails to form and the hyaloid vasculature persists,

16

continuing to supply the mature retina. These abnormal vessels typically invade
and neovascularize the neural retina, and they are prone to hemorrhage, which
can occur in the subretinal space, vitreous, or spread into the anterior chamber
(Brown et al., 2001; Brzezinski et al., 2003).
Diseases of the optic nerve and retinal vasculature
Congenital optic nerve diseases are important causes of hereditary
blindness (Taylor, 2007). Optic nerve hypoplasia (ONH), in which nerves are
reduced in size, and optic nerve aplasia (ONA), in which nerves are completely
absent, are thought to be caused by primary defects in the specification,
differentiation, or survival of RGCs (Lambert et al., 1987). The less severe
clinical entity, ONH, is much more common than ONA (Borchert and GarciaFilion, 2008). Each can occur as an isolated malformation, or as part of a
syndrome, together with central nervous system (CNS) and pituitary defects
(Borchert and Garcia-Filion, 2008; McCabe et al., 2011). In ONA, the mature
retinal vasculature typically fails to develop, and fetal vessels may persist in the
vitreous (Blanco et al., 1992; Brodsky et al., 2004; Lee et al., 1996; Little et al.,
1976; Scott et al., 1997). Most cases of ONA and ONH are sporadic. Few
genetic causes have been identified and these explain only a very small fraction
of cases. Mutations in HESX1, OTX2, SOX2 or PAX6 can lead to severe ONH
or ONA (Azuma et al., 2003; Dattani et al., 1998; Kelberman and Dattani, 2007;
McCabe et al., 2011; Ragge et al., 2005). These Mendelian cases usually
present with hypothalamic-pituitary dysfunction, and global eye and CNS
malformations. The diagnostic terms Syndrome of Optic Nerve Hypoplasia

17

(SONH), Septo-Optic Dysplasia (SOD) or de Morsier syndrome (MIM 182230)


are applied when ONH coexists with hypoplasia of the midline structures such as
the pituitary, hypothalamus or septum pellucidum (Borchert and Garcia-Filion,
2008; Brodsky and Glasier, 1993; De Morsier, 1962; Haddad and Eugster, 2005).
Other hereditary diseases may affect the physiology of ganglion cells.
These include Lebers hereditary optic neuropathy and autosomal dominant optic
atrophy. These diseases involve mtDNA genes or autosomal loci that encode
mitrochondrial proteins. The pathology is predominantly caused by mitochrondrial
dysfunction, which ultimately leads to RGC cell death (Abu-Amero, 2011; Olichon
et al., 2006). It remains to be determined why RGCs are particularly prone to
these metabolic defects, but it is hypothesized that accumulation of reactive
oxygen species in mitochrondria at the optic disc leads to degeneration (Ghelli et
al., 2003; Howell, 1998; Olichon et al., 2006) and may be correlated with the
degree of myelination (Carelli et al., 2002; Sadun and Carelli, 2003).
Glaucoma is a leading cause of bilateral blindness worldwide (Lee and
Higginbotham, 2005; Quigley, 1996). Genetic analysis has provided many
insights into the pathogenesis of this disease (Allingham et al., 2009; Challa,
2004). Glaucoma is caused by optic nerve head injury or dysfunction, which
ultimately leads to RGC cell death. Often this insult is triggered by increased
intraocular pressure (IOP) due to inadequate drainage of front of the eye (the
anterior segment). However, many cases of glaucoma can present with normal
IOP (normal tension glaucoma), and thus may be caused by primary defects in
RGC function (Buono et al., 2002), and longstanding ocular hypertension has

18

been noted in some patients with normal vision (Copt et al., 1999). Despite
numerous advances in recent years, many cases of glaucoma remain
unexplained, and the pathogenesis is incompletely understood.
Retinal vascular diseases are also major causes of blindness (Gariano
and Gardner, 2005). Retinopathy of prematurity (ROP) is vascular disorder that
primarily affects premature infants and can be triggered by excessive oxygen
exposure in newborns. It is characterized by vascular loss and incomplete
vascularization followed by a compensatory hyperproliferation of the remaining
retinal vessels (Chen and Smith, 2007). ROP can be accompanied by persistent
hyperplasia of the primary vitreous (PHPV). In this disorder, the primary fetal
vasculature fail to regress, likely due to impaired apoptosis, abnormalities in the
mature vascular development, or hypoxic conditions (Goldberg, 1997; Reese,
1955; Shastry, 2009). Most cases are sporadic, but rare familial cases have
been characterized (Haddad et al., 1978; Khaliq et al., 2001). PHPV predisposes
retinas towards detachment, hemorrhage, and anterior chamber defects (Pruett,
1975). A small number of cases may be explained by genetic defects in the Wnt
signaling pathway (Robitaille et al., 2009). Mutations in these genes can also
cause Norries disease or familial exudative vitreoretinopathy (FEVR), which
share similar clinical features (Berger et al., 1992a; Berger et al., 1992b; Gariano
and Gardner, 2005; Robitaille et al., 2009). Few other genetic causes for these
diseases have been identified.

19

Novel insights into RGC development and the role of Math5 (ATOH7)
Ganglion cells play an essential role in the visual system. They transmit
all visual information to the CNS and are required for the development of retinal
vasculature. Thus, the development of RGCs remains an important area of study
for understanding the causes of blindness and the mechanisms of visual system
development. In this dissertation, I have focused my analysis on the ATOH7
(Math5) gene, which plays a particularly important role in the initial development
of this retinal cell type.
In Chapter II, I thoroughly and systematically analyze the transcriptional
architecture of the mouse Math5 gene. I found that this single-exon gene makes
one predominant mRNA isoform, which is unspliced. This study starkly contrasts
a prior report that suggested a major role of alternative splicing in the regulation
of the gene (Kanadia and Cepko, 2010).
In Chapter III, Joseph Brzezinski and I quantitatively characterized the
lineage and cell cycle properties of the Math5-expressing cells. We show that
Math5 is variably expressed in progenitors during or after their terminal cell cycle,
with a gradual restriction to post-mitotic cells as development proceeds. Using
Math5>Cre BAC transgenic mice, we show that only 11% of lineage-derived cells
adopt the RGC fate, and that only 55% of ganglion cells are marked by Math5
expression. Surprisingly, the Math5-independent fraction is substantial even
during early development, suggesting a non-autonomous role of Math5 in
ganglion cell fate specification.

20

In Chapter IV, I extend the analysis of Math5 cell cycle dynamics by


examining the properties of key transcription factors in RGC, amacrine, and
horizontal cell development. Surprisingly, I found that Brn3b, a marker of
committed RGCs, and co-regulator Isl1 can be expressed in some cells prior to
the last progenitor mitosis. Consistent with fate specification prior to cycle exit, I
observed pairs of Brn3b+ cells during cytokinesis and clonal pairs of RGCs
resulting from terminal symmetric divisions in retinal explants. These results
suggest that the cell fate decision is not synchronized to cycle exit.
In Chapter V, I thoroughly test the sufficiency of Math5 (Atoh7) in
promoting ganglion cell fate, or biasing retinal progenitors. By broadly
overexpressing Math5 in mouse neurogenic cells using a transgene with Crx
(cone-rod homeodomain) regulatory DNA, I found that Math5 does not
extensively promote RGC fate in the wild-type retinal environment. In contrast,
ectopic Math5 can partially rescue RGC fate in Math5 -/- mice. However, given
the late temporal profile of transgene expression, the rescue is heterochronic,
such that early-born RGCs are absent. The transgene-derived, late-born RGCs
exhibit pathfinding defects and undergo apoptosis, suggesting that early RGCs
provide vital trophic support for late-born RGCs.
In Chapter VI, we explore the role of human ATOH7 in optic nerve and
vascular diseases of the retina. We identified point mutations in a family with
autosomal recessive PHPV and a sporadic case of bilateral optic nerve
hypoplasia. Subsequently, I characterized these point mutations through
biochemical and functional studies. I found a p.N46>H mutation in the arPHPV

21

family to be the likely cause of disease, as the mutant protein is incapable of


binding to DNA, activating transcription, or rescuing ganglion cell development in
Atoh7 -/- mouse retinal explants. Taken together, this dissertation highlights the
importance of ATOH7 in eye development and disease, and provides novel
insights into the structure and function of ATOH7 and the development of RGCs.

22

Figure I-1. Structure of the eye and retina. (A) Basic fuschin and methylene
blue staining of a central section of a mouse eye. Major structures of the eye
are highlighted. (B) H+E staining showing the trilaminated mouse retina. A
schematic diagram of the 7 major retinal cell types is shown on the right. rpe,
retinal pigmented epithelium; on, optic nerve; ret, retina; cb, ciliary body; corn,
cornea; onl, outer nuclear layer, inl, inner nuclear layer; gcl, ganglion cell
layer; ilm, inner limiting membrane; olm, outer limiting membrane; r, rod; c,
cone; hz, horizontal neuron; bip, bipolar cell; am, amacrine; da, displaced
amacrine; rgc, retinal ganglion cell; mg, Mller glia.

23

Figure I-2. Molecular mechanisms of retinal development. (A) Schematic


birthdating curves for the 7 major cell types. Ganglion cells are the first born celltype, but there is extensive overlap among birthdates across the 7 cell types.
Adapted from (Marquardt and Gruss, 2002). (B) The serial competence model
posits that retinal progenitors pass through discrete competence states in which
they can adopt a small number of cell fates. Cell fate within these states is
determined largely by environmental influences. Adapted from (Livesey and
Cepko, 2001). (C) Adapted from (D) Transcription factors in retinal
development. The diagram shows the major steps in retinal histogenesis, and
several key intrinsic transcription factors that are involved in the specification and
differentiation of each cell type.

24

25

Figure I-3. Math5 (Atoh7) knockout mice have no optic nerves, and
severe deficiencies in RGCs. (A-B) Views of the brain of wild-type (A)
and Math5 -/- mice (B). The optic chiasm and nerves are absent in
Math5 -/- mice. (C-D) Hematoxylin and eosin staining of retinal sections
from wild-type (C) and Math5 -/- (D) mice. Math5 mutant retinas are
thinner than wild-type, and have a hypocellular ganglion cell layer that is
composed of displaced amacrine cells. Adapted from Brown et al., 2001.
onl, outer nuclear layer; inl, inner nuclear layer; gcl, ganglion cell layer.

26

Figure I-4. Development of the retinal vasculature. (A-H) Vasculature


development in wild-type (WT) and Math5 -/- mice. In wild-type mice (A, C, E,
G), intrinsic retinal vasculature (red, arrows) spreads radially following an
astrocyte network, while hyaloids vessels (arrowheads) regress over time. In
Math5 -/- mice (B, D, F, H), intrinsic vasculature fails to sprout by angiogenesis
from the area of the putative optic nerve (*). Instead, hyaloid vessels proliferate
and spread across the inner retina. (I) Models of vascular development in wildtype and Math5 -/- mice. Images adapted from (Brzezinski, 2005) and (Edwards
et al., 2011).

27

28

CHAPTER II
A CRITICAL ANALYSIS OF MATH5 (ATOH7) mRNA SPLICING IN THE
DEVELOPING MOUSE RETINA

Abstract
The Math5 (Atoh7) gene is transiently expressed during retinogenesis by
progenitors exiting mitosis, and is essential for ganglion cell (RGC) development.
Math5 contains a single exon, and its 1.7 kb mRNA encodes a 149-aa
polypeptide (Brown et al. 1998). Mouse Math5 mutants have essentially no
RGCs or optic nerves. Given the importance of this gene in retinal development,
we thoroughly investigated the possibility of Math5 mRNA splicing by Northern
blot, 3 RACE, RNase protection assays, and RT-PCR, using RNAs extracted
from embryonic eyes and adult cerebellum, or transcribed in vitro from cDNA
clones. Because Math5 mRNA contains an elevated G+C content, we used
graded concentrations of betaine, an isostabilizing agent that disrupts secondary
structure. Although ~10% of cerebellar Math5 RNAs are spliced, truncating the
polypeptide, our results show few, if any, spliced Math5 transcripts exist in the
developing retina (<1%). Rare deleted cDNAs do arise via RT-mediated RNA
template switching in vitro, and are selectively amplified during PCR. These data
differ starkly from a recent study (Kanadia and Cepko 2010), which concluded
that the vast majority of Math5 and other bHLH transcripts are spliced to

29

generate noncoding RNAs. Our findings clarify the architecture of the Math5
gene and its mechanism of action.

Introduction
The vertebrate retina develops from a single multipotent progenitor
population, which gives rise to seven major cell types rod and cone
photoreceptors; amacrine, bipolar and horizontal interneurons; Muller glia; and
retinal ganglion cells (RGCs) (Holt et al., 1988; Turner and Cepko, 1987). These
diverse cell types emerge from the mitotic progenitor pool in rough sequential
order, with overlapping birthdates (Livesey and Cepko, 2001; Wong and
Rapaport, 2009). RGCs are the first-born retinal cell type in every vertebrate
examined (Altschuler et al., 1991). These cells transmit all visual information from
the eye to the brain, via their axons, which comprise the optic nerves. The gene
network regulating retinogenesis is an active area of investigation.
An important clue toward understanding the mechanism of vertebrate
retinal fate specification was the discovery of Math5 (Atoh7), a proneural basicloop-helix (bHLH) transcription factor that is evolutionarily related to Drosophila
Atonal and mouse Math1 (Atoh1) (Brown et al., 1998; Kanekar et al., 1997). The
mouse Math5 gene is expressed transiently in retinal cells exiting mitosis, from
E11.5 until P0, in a pattern that is correlated with the onset of neurogenesis, and
it is necessary for RGC fate specification. Math5 mutant mice lack RGCs and
optic nerves (Brown et al., 2001; Wang et al., 2001), and have secondary defects
in retinal vascularization (Brzezinski et al., 2003) and circadian photoentrainment

30

(Brzezinski et al., 2005). In zebrafish, the homologous lakritz mutation also


causes RGC agenesis (Kay et al., 2001), and in humans, the ATOH7 gene may
be associated with congenital optic nerve disease (Brown et al., 2002). Although
the exact mechanism of Math5 action remains unknown, it is thought to confer an
RGC competence state on early retinal precursors (Brzezinski and Glaser, 2004;
Yang et al., 2003). A number of potential target genes are misregulated in Math5
mutant retinas (Mu et al., 2005). Apart from the retina, expression domains have
been defined in the hindbrain cochlear nucleus and cerebellum (Saul et al.,
2008).
During our initial characterization of Math5 (Brown et al., 1998), we
identified multiple independent retinal cDNA clones, which were colinear and
coextensive with mouse genomic DNA. The internal sequence and termini of
these clones were consistent with a single-exon transcription unit.
In a recent provocative study, Kanadia and Cepko (2010) report that the
vast majority of Math5 transcripts in embryonic mouse retinas are spliced, with
donor and acceptor sites located in the 5 and 3 UTRs, such that the coding
sequences are excised. This conclusion, which plainly differs from our previous
studies (Brown et al., 2002; Brown et al., 1998), was based largely on the size
and abundance of particular RT-PCR products. Similar observations were
reported for Ngn3 (neurogenin, Neurog3), a related bHLH factor. If correct, these
findings raise important questions regarding the origin, extent and function of
noncoding (nc) bHLH-gene RNAs, which may integrate into larger gene
regulatory networks during neural development (Mercer et al., 2009), and

31

suggest that abortive splicing may be utilized as a novel post-transcriptional


mechanism to regulate bHLH gene expression. Given the importance of Math5
for retinogenesis, the central role of bHLH factors in neuronal fate specification
(Bertrand et al., 2002), and the possibility that functional coding and noncoding
RNAs may be generated in the same orientation by alternative splicing of a
single transcription unit (Chooniedass-Kothari et al., 2004), we have
systematically evaluated Math5 mRNA splicing in the developing retina, using
RNA hybridization and RT-PCR methods adapted for the extreme G+C content
of the transcript.
Our data strongly suggest that the apparently frequent splicing of Math5
retinal mRNA is a technical artifact, resulting from: (1) profound secondary
structure in the mRNA, promoting template switching during reverse transcription
in vitro, (2) selective amplification of deleted products lacking the internal GC-rich
segment; and (3) the existence of very rare mis-spliced molecules, representing
less than one percent of Math5 transcripts. Our results refine the structure of the
Math5 transcription unit, explore the concept of an intronless gene, and provide a
cautionary lesson for PCR-based studies of RNA processing.
Materials and Methods
Plasmid clones and oligonucleotides
Math5 clone pJN4C (accession nos. AF071223, AF418923) was derived from a
neonatal C57BL/6 retinal cDNA library (Brown et al., 1998). It contains 318 bp
5 UTR, 447 bp coding sequence (CDS) and 390 bp 3 UTR, and terminates at an
A-rich stretch in the 3 UTR. Clones JN1 and JN2 extend 55 bp and 279 bp further

32

in the 5 and 3 directions, respectively (Fig. II-1). Plasmid vector pCR4-TOPO


(Invitrogen) was used for TA cloning of RT-PCR products, including the
templates used for RPA probes. All custom PCR primers in this study are
indicated in Fig. II-1A and listed in Table II-S1.
RNA
Total RNA was isolated from eyes or retinas dissected from CD-1 mouse
embryos (ages E14.5 and E15.5) and adult tissues (eyes, cerebral cortex,
cerebellum and liver) by the phenol-guanidinium-chloroform (Trizol) extraction
method (MacDonald et al., 1987).
Northern analysis
Ten g total RNA from each tissue was resolved by formaldehyde-agarose gel
electrophoresis and transferred to a 0.45 m pore nitrocellulose membrane as
described (Cho and Dressler, 1998). An RNA ladder in the 0.25-9.5 kb range
(Gibco-BRL) was co-electrophoresed to accurately determine the size of
hybridizing RNAs. After prehybridization, the membrane was probed
successively with 32P-radiolabeled 1.2 kb Math5 and 1.1 kb -actin (Alonso et al.,
1986) mouse cDNAs, washed to 0.1X SSC 65C stringency, and exposed to
Kodak XAR film with an intensifying screen at -80C for 16 hrs. The
autoradiographic images were digitized using a flatbed scanner. The Math5
probe was gel-purified from clone JN4C after digestion with XhoI and EcoRI, and
was labeled to high specific activity with 32P-[]-dCTP using the random hexamer
(dN6) priming method (Feinberg and Vogelstein, 1983).

33

Reverse transcriptase (RT) and genomic PCRs


Total RNA from E14.5 or E15.5 mouse eyes (5 g) or retinas (3 g), adult
cerebellum (5 g), or adult liver (3 g) was treated with 5 U DNaseI (Roche) for
15 min at 37C in DNAse buffer (20 mM Tris-HCl, 2 mM MgCl2, 50 mM KCl). To
stop the reaction, EDTA was added to 2 mM and the DNaseI was inactivated at
75C for 10 min. RNAs were mixed with 500 ng oligo dT or 300 ng dN6
(Invitrogen) primer, denatured at 65C for 10 min, and reverse-transcribed with
10 U TranscriptorTM High Fidelity RT (Roche) at 55C for 1 hr. The 20 L RT
reactions contained 50 mM Tris-HCl pH 8.5, 30 mM KCl, 8 mM MgCl2, 5 mM DTT,
1 mM dNTPs, and 10 U RNase Inhibitor (ProtectorTM, Roche). The RT was
inactivated at 85C for 5 min. The TranscriptorTM enzyme mixture has RNAdirected DNA polymerase, DNA-dependent DNA polymerase, helicase,
RNase H, and 3 5 exonuclease proofreading activities (Schonbrunner et al.,
2006). In the RT (-) controls, this enzyme mixture was replaced with nucleasefree water.
PCRs were performed using 1 L of the cDNA reactions as template, in
1.5 mM MgCl2, 0.2 mM dNTPs, 20 mM Tris pH 8.4, 50 mM KCl, with 2 nM each
primer and 2.5 U hot-start Platinum Taq polymerase or 0.5 U conventional Taq
polymerase (Invitrogen). All PCRs were performed in 12-well strip tubes, in a
96-well MJ thermocycler with heated lid assembly, using specified primers and
conditions (Table II-S1, 2). PCR products were separated by electrophoresis
through 1.5 % agarose gels, purified by membrane binding (Wizard SV,

34

Promega) and sequenced or subcloned. Genomic PCRs were performed using


50 ng CD-1 mouse tail DNA.
To melt secondary structure, 10X MasterampTM (Epicentre) was included
in some PCRs, with a final fractional volume in the reaction mixture between 0.0
to 0.3 (v/v), designated 0X to 3X. Although the formulation of this additive is
proprietary, equivalent results were obtained with 0.0 to 1.0 M betaine (Sigma
B0300).
In vitro transcription
Plasmid DNA (1 g) from clones pJN4C (Brown et al., 1998) or pCR4-ECO was
linearized with XhoI or NotI, respectively, and transcribed for 2 hr with 40 U
bacteriophage T3 RNA polymerase (Roche), in a reaction containing 1 mM
NTPs, 40 mM Tris-HCl pH 8.0, 6 mM MgCl2, 10 mM DTT, 2 mM spermidine and
20 U RNase Inhibitor (Protector). The template was then digested with 20 U
DNase I for 1 hr at 37C, and the resulting RNA was purified using Trizol
(Invitrogen) and assessed by 1% agarose gel electrophoresis and UV
absorbance (A260). The full length (FL) Math5 IVT RNA product (10 ng) was
mixed with DNaseI-treated mouse liver RNA (3 g) or used directly (10-200 ng)
for RT-PCRs.
Triplex competitive RT-PCR assays
Retinal and cerebellar RT-PCRs (Leygue et al., 1996) were performed in 1X
MasterAmpTM, with equal molar ratios of competing forward primers (1 nM) and a
single fluorescent (6-FAM) reverse primer (LP4) as indicated (Table II-S2), which

35

were matched for length and G+C content. Products were diluted to 1:50 to 1:200
in formamide and co-electrophoresed with GS-600 LIZ size marker in a 3730XL
capillary DNA Analyzer (Applied Biosystems). The fluorescence intensity of each
amplimer and the ratio of spliced to unspliced PCR products were calculated
using GeneMarker (SoftGenetics), from the sum of major peaks in triplicate
experiments.
Rapid amplification of cDNA ends (3 RACE)
First-strand cDNA synthesis was performed from retinal RNA as described
above, using 10 pmol adapter primer (AP, Table II-S1). One L of the cDNA
reaction was then used to amplify 3 terminal sequences using primers and
conditions in Tables II-S1 and II-S2. To minimize spurious products from
unrelated genes, a second round of PCR was performed using nested primers,
following a conventional nested 3 RACE strategy (Frohman, 1993).
RNase protection assays (RPA)
RNase protection assays (Melton et al., 1984) were conducted using the RP-III
kit (Ambion). Antisense cRNA probes were transcribed from PCR products A
and B (Fig. II-5) cloned in pCR4-TOPO. One g of each plasmid was digested
with NotI and transcribed with T3 RNA polymerase as described above, except
that 125 pmol (75 Ci) or 113 pmol (90 Ci) 32P-[]-CTP was included in probe A
and B reactions, respectively, with 200 pmol CTP and 10 nmol of ATP, GTP and
TTP. This yielded 366 and 632 nt cRNA products with 301 nt (A) and 567 nt (B)
direct sequence homology to Math5. Probes were purified by electrophoresis

36

through denaturing 6% polyacrylamide gels and eluted for 3-4 hrs at 37C. Ten
g of DNaseI-treated E14.5 eye RNA, yeast RNA (Ambion), or yeast RNA spiked
with 10 ng Math5 IVT product (ECO or FL) was precipitated in 2.5 M ammonium
acetate 70% ethanol and resuspended in 8 l hybridization buffer. The RNAs
were hybridized with 2 l probe A (8 x 104 cpm) or probe B (1.2 x 105 cpm) for
13 hr at 42C, and digested with RNase A+T1 (1:100) for 30min at 37C, and
co-precipitated with glycogen and 5 g yeast carrier RNA. Reactions were
electrophoresed through 6% polyacrylamide denaturing gels (0.4 mm) in 6 M
Urea and 0.5X TBE. dsDNA size markers were prepared by radiolabeling MspIdigested pBR322 with 32P-[]-dCTP and Klenow DNA polymerase. The dried
gels were exposed to phosphor screens for 12-14 hrs and imaged using a
Typhoon scanner (Molecular Dynamics) at 0.2 mm resolution. Yeast RNA
controls were included RNase, to assess the probe integrity and the
completeness of digestion.
Informatics
Sequence alignments, G+C and antigenicity profiles, and PCR primer
optimization were performed using MacVector (Accelrys) software and NCBI
BLAST servers. Math5 polyadenylation sites were predicted using the polyADQ
(Tabaska and Zhang, 1999) web server (rulai.cshl.org/tools/polyadq/). Scores
were calculated for a 6.0 kb sequence extending from the transcription start site
(Fig. II-1D), using default threshold values. RNA secondary structures were
predicted by free-energy minimization (Mathews et al., 1999; Zuker, 2003) using
the M-fold web server (mfold.bioinfo.rpi.edu/). Expressed sequence tags (ESTs)
37

for mouse and human bHLH cDNAs were accessed through the UCSC genome
browser (genome.ucsc.edu/).
Math5 antibodies
Commercial and custom antibodies to Math5 peptides and recombinant proteins
are indicated in Fig. II-S2 along with the immunogen, including the Abcam
polyclonal reagent (ab13536) cited by Kanadia and Cepko (2010). Custom rabbit
polyclonal sera were generated using internal (RCEQRGRDHP) or C-terminal
(RLFGFQPEPFPMAS) Math5 peptide haptens coupled to KLH (keyhole limpet
hemocyanin) via a cysteine thiol linkage (Research Genetics, Huntsville, AL), and
were affinity purified.
Cell transfection and Western analysis
NIH 3T3 fibroblast cultures were transfected with expression plasmid DNA (1 g
per 60 mm plate) for native or N-terminal 6x Myc-tagged versions of mouse or
human ATOH7 proteins, or empty vector, using Fugene-6 reagent (Roche), with
0.1 g pUS2-EGFP as an internal control. These plasmids were prepared by
inserting ATOH7 coding regions from genomic phage, plasmid or BAC clones
into pCS2 and pCS2MT vectors (Rupp et al., 1994) and verifying the sequence.
Mouse pCS2-Math5 and pCS2MT-Math5 plasmids were described previously
(Brown et al., 1998). After 48 hrs, cells were harvested in PBS with protease
inhibitors (CompleteTM, Roche), lysed in RIPA buffer (Harlow and Lane, 1988),
sonicated, and centrifuged at 13,000 g for 15 min at 4C. Soluble proteins were
electrophoresed through NuPAGE Novex Bis-Tris 4-12 % polyacrylamide gels

38

(25 g per lane), transferred to nitrocellulose membranes and stained with


Ponceau S. Parallel Western blots were probed with rabbit polyclonal antisera to
Math5 peptides (1:200, Fig. II-S2A), full-length human ATOH7 (D01P, 1:500) or
GFP (Abcam ab290, 1:2500); or mouse anti-Myc monoclonal (9E10, Zymed,
1:500); and the reactive proteins were visualized using HRP-conjugated antirabbit (NEN, 1:5000) or mouse (GE, 1:20,000) IgG secondary antibodies,
enhanced chemiluminesence reagents (ECL-Plus, GE), and Kodak MS X-ray
film.
Immunostaining and RNA in situ hybridization
Mouse E15.5 embryo heads from wild-type and Math5 knockout (Atoh7 tm1Gla)
littermates (Brown et al., 2001) were fixed in 4% paraformaldehyde PBS for 1 hr
at 4OC; processed through a 10-30% sucrose series in PBS; cryoembedded in
OCT media (Tissue-Tek, Torrence, CA) and sectioned through the eyes at
5-10 m. To thoroughly test antibody reactivity, we tried three different antigen
unmasking protocols in parallel: 0.1 M Tris pH 9.5 at 95C for 5 min; 0.05 %
trypsin at 37C for 10 min; and 0.3% Triton X-100 0.1 M Tris pH 7.4 at 25C for
10 min. Cryosections were then blocked and processed in TST milk as described
(Mastick and Andrews, 2001). Slides were incubated overnight at 25C with a
1:500 dilution of anti-Math5 peptide sera (Abcam no. ab13536, lot no. 610696),
followed by a 1:5000 dilution of Alexa 594-conjugated goat anti-rabbit IgG
secondary antibody (Molecular Probes). RNA in situ hybridization was performed
on E15.5 embryonic retinas as described (Wallace and Raff, 1999). A
digoxigenin-labeled antisense Math5 cRNA probe spanning the 3 UTR and CDS
39

was prepared from AscI-digested plasmid pJN4C with T7 RNA polymerase,


hybridized to retinal sections overnight, detected using an AP-conjugated sheep
anti-DIG antibody (1:2000, Roche), and visualized using NBT-BCIP
histochemistry. Micrographs were imaged using a Zeiss Axioplan 2 microscope,
digital camera and Axiovision software.

Results

Math5 transcription unit, defined by cDNA clones, Northern and 3 RACE


analysis
During our initial characterization of Math5 (Brown et al., 1998), we
identified four independent retinal cDNA clones, which were colinear with mouse
genomic DNA (Genbank accession no. AF418923). The 5 and 3 termini, and
internal sequences were consistent with RNA hybridization data suggesting a
single-exon transcription unit, with an initiation site 23 bp downstream from a
TATAAA box and a polyadenylation (pA) site 669 bp downstream from the TAA
stop codon, giving 1.7 kb as the predicted size for polyA+ Math5 mRNA (Fig. II1A,D). This major Math5 transcript was detected by Northern blot analysis of
E15.5 mRNA with an 1155 bp radiolabeled cDNA probe (JN4C) that includes
318 bp 5 UTR, 447 bp coding sequence (CDS) and 390 bp 3 UTR (Fig. II-2A). A
second, less abundant 4.4 kb transcript was also detected at this age, which is
close to the peak time-point for Math5 expression during embryogenesis
(Brzezinski and Glaser, 2004). Careful inspection of the autoradiogram, in
relation to the RNA size markers, revealed no smaller Math5 transcripts,

40

particularly in the 0.8-1.0 kb size range expected for spliced isoforms lacking the
coding region. This pattern resembles Northern data obtained by Kanadia and
Cepko with UTR probes (cf. Fig.1f and 1f), but appears inverted compared to the
unsized blot hybridized with a CDS probe in their report (cf. Fig. 1f). We cannot
explain this discrepancy.
To confirm our identification of the major Math5 polyadenylation site
(Brown et al., 1998) and define the 3 terminus of the longer, 4.4 kb transcript, we
first surveyed the 3 Math5 genomic region for favorable pA signals using the
polyADQ weighted statistical algorithm (Tabaska and Zhang, 1999). Among eight
potential pA sites downstream from the transcription start site (TSS), two had
significant polyADQ scores (nos.1 and 6, Fig. II-2B,C), and these were consistent
with the observed transcript sizes. We then looked for mRNAs terminating at pA1
and pA6 in parallel 3 RACE experiments (Frohman, 1993), using E14.5 total eye
RNA and nested primers positioned upstream of each site (Figure 2b,d). From
the size and sequence of the products (Fig. II-2D-F), and our Northern data, we
conclude that there are two principal Math5 transcripts in the retina, 1.7 kb and
4.4 kb in length, and that both of these transcripts are unspliced. This
interpretation is further supported by the curation of additional mouse cDNAs,
represented as 56 expressed sequenced tags (ESTs) and two Genbank cDNAs
in the NCBI database (Fig. II-S1). Only two ESTs and one cDNA, originating from
the adult cerebellum, appear to be authentic splice products (see below), and
these do not correspond to the retinal isoforms reported by Kanadia and Cepko
(2010).

41

In addition to the coding region, Math5 mRNA has three notable features
relevant to this study (Fig. II-1A). First, the 5 half is highly enriched in G+C
nucleotides (Fig. II-1B), with >85% G+C content in the 150 nt segment spanning
codons 7 to 57. Math5 mRNA thus has the potential to form compact,
thermodynamically stable secondary structures, owing to the third hydrogen bond
in G-C pairs compared to A-U pairs, and the ability of guanine residues to
interact with uracil in folded RNA (Mathews et al., 1999). The elevated G+C
content is also predicted to affect folding of the (+) and (-) strand cDNA
templates, compromising DNA polymerase processivity. Second, the 5 segment
of the gene is enriched for specific trinucleotide elements (Py-G-C) that are
known to cause DNA polymerase pausing (Mytelka and Chamberlin, 1996) (Fig.
II-1C). These account for 15.7% of the trinucleotides in this segment (47 of 300,
for both DNA strands), which is 1.73 fold higher than expected from
mononucleotide frequencies. Third, mouse Math5 mRNA contains 30-nucleotide
imperfect direct repeats (DRs), located in the 5 and 3 UTRs (Fig. II-1A,E).
These UTR repeats are not conserved among mammalian ATOH7 mRNAs.
Sensitivity of Math5 PCR to template folding in vitro
Our Northern analysis, screening of cDNA libraries, and analysis of ESTs
contrasts starkly with the abundant, heterogeneous splicing recently reported for
the Math5 gene (Kanadia and Cepko, 2010). As a first step to resolve this
difference, we performed a series of RT-PCR experiments using the same
primers (LP8 and LP4, Fig. II-1A and Table II-S1) and similar conditions (Table IIS2) as these authors. Using a thermostable reverse transcriptase (RT)

42

formulation (TranscriptorTM, Roche), E14.5 total mouse retinal RNA as template,


and primers located in the 5 and 3 UTRs, we amplified a single 448 bp product
(Fig. II-3A) with the same sequence as the ECO cDNA reported by Kanadia and
Cepko (2010) (Fig. II-3C), thus technically reproducing their primary observation.
In this cDNA, a 639 bp segment encompassing the entire Math5 coding region
has been deleted. The 3 breakpoint abuts the 3 UTR direct repeat.
The extremely high G+C content of the 5 half of the deleted segment (Fig.
II-1B) creates the potential for the RNA to form stable secondary structures,
which could impede the procession of reverse transcriptase (RT) and DNA
polymerases. Given our previous experience working with Math5, we repeated
this PCR, replacing the water in the reaction mixture with 0 to 3X MasterampTM
(Epicentre). This is functionally equivalent to 0 to 1.0 M betaine (N,N,N-trimethyl
glycine) [not shown], which is the principal ingredient in this additive (Henke et
al., 1997; Mytelka and Chamberlin, 1996; Weissensteiner and Lanchbury, 1996).
In these reactions, betaine interacts with DNA as an isostabilizing agent,
equalizing the free energies of A-T and G-C pairs by increasing hydration of the
minor groove and flexibility of the double helix (Melchior and Von Hippel, 1973;
Rees et al., 1993). It thus melts secondary structures, allowing DNA polymerases
to extend through GC-rich segments (Henke et al., 1997; Mytelka and
Chamberlin, 1996; Weissensteiner and Lanchbury, 1996). In our experience,
1 M betaine is required to reliably amplify across the 5 coding sequences of
mouse or human ATOH7, even when cloned cDNA is used as a template; and
relatively high concentrations (~2 M) are tolerated in the PCR. Moreover, in the

43

absence of betaine, we have observed numerous PCR-generated deletions of


Math5 sequences during molecular cloning projects over several years (not
shown).
As the concentration of betaine in the PCR was increased, the apparently
spliced 448 bp ECO product vanished, and a strong 1087 bp product appeared,
corresponding to full-length, unspliced Math5 cDNA (Fig. II-3A). The identity of
these molecules was verified by sequencing gel-purified PCR products and
multiple pCR4-TOPO plasmid clones derived from the PCR products. The effect
of betaine on the generation of the 448 bp product suggests that Math5 splicing
either does not occur in nature, within the developing retina, or is an extremely
rare event. Indeed, under normal circumstances, the smaller product should have
been significantly favored during the amplification steps, with or without betaine.
However, since the ECO product cannot be generated by PCR from mouse
genomic DNA (Fig. II-3B) (Melchior and Von Hippel, 1973; Rees et al., 1993) and
depends on RT, it must be represented in the initial first-strand cDNA pool, albeit
at an extremely low level (see below). These molecules could have been
generated from rogue, aberrantly spliced mRNAs or by RNA template-switching
during the reverse transcription step. Regardless of their origin, these rare cDNA
amplicons (448 bp, 52.7 % GC) should have a large selective advantage over the
full-length co-terminal cDNA (1087 bp, 60.1 % GC) during subsequent cycles of
PCR.
Similar experiments were performed with a second pair of primers (LP6
and LP7), which are separated by 486 bp in genomic DNA and flank the GC-rich

44

segment (Fig. II-3C,D). In the absence of betaine, these primers did not amplify
any product. However, when 2-3X MasterampTM was included in the PCR, only
the expected 486 bp amplimer was observed. When we extended the PCR
beyond 35 cycles, preincubated the reaction at 25C (cold start) or used crude
Taq polymerase preparations in the absence of betaine, a heterogeneous group
of deleted (lacunar) products was observed (not shown), with a size and
sequence distribution (Fig. II-4D, Table II-S3) similar to that reported by Kanadia
and Cepko.

Deleted PCR products derived from RNAs transcribed in vitro


To determine the origin of the lacunar cDNAs, we performed parallel
RT-PCR experiments on RNA templates derived by in vitro transcription (IVT).
Full-length, sense Math5 transcripts were synthesized in vitro using
bacteriophage T3 RNA polymerase and a XhoI-cleaved pJN4C DNA template
(Fig. II-4A). RT reactions were performed as before, with oligo dT-priming, and
0-200 ng of the in vitro RNA transcript as template, alone or diluted into 3 g total
mouse liver RNA. When the PCRs were performed in 3X MasterampTM,
full-length 1087 bp and 486 bp products were amplified (Fig. II-4B,C), identical to
those generated from E14.5 retinal RNA (Fig. II-3A,C). However, when the
betaine was reduced or omitted, we observed a variety of smaller products, with
a size distribution (Fig. II-4B-D) and sequence diversity (Table II-S3) similar to
that reported by Kanadia and Cepko (2010, cf. Table S1), despite the absence of
retinal RNA, spliceosomes or other eukaryotic cell components.

45

Because these products depend on reverse transcriptase, they must have


arisen via RNA template-switching during the RT reaction (Biagini et al., 2008),
despite the use of a thermostable recombinant enzyme mixture with high fidelity,
processivity and proofreading features (Kitabayashi and Esaka, 2003;
Schonbrunner et al., 2006). A similar origin seems likely for the majority of
apparently spliced Math5 cDNAs reported by Kanadia and Cepko (cf. Table S1).
Indeed, most of the deleted products obtained here and in the previous paper
(Table II-S3) contain 5-10 nt direct sequence homology at the junctions (Pfeiffer
and Telesnitsky, 2001). The only remaining explanation that Math5 encodes a
nuclear self-splicing mRNA lacks precedent (Cech, 1986). A possible exception
is the ECO cDNA product, which was amplified from embryonic retinal RNA in
less than 1M betaine (Fig. II-3A) but not from IVT-derived material or genomic
DNA.
Critical evaluation of Math5 splicing by competitive RT-PCR
To further investigate Math5 splicing in vivo, we directly compared the
abundance of full-length and ECO (spliced) RNAs in competitive, triplex
(3-primer) RT-PCR assays (Fig. II-5). Each reaction contained two alternative
forward (sense strand) primers one located in the 5 UTR and a second,
internal primer in the 3 coding region plus a single reverse (antisense) primer
located in the 3 UTR (Fig. II-5A). In this assay, the proportion of the two
predicted products should reflect the relative abundance of the corresponding
mRNAs in the E14.5 retina. The outer UTR primers and the resulting ECO
product are identical to those reported by Kanadia and Cepko (see Table II-S1).

46

For completeness, we performed two independent competitive RT-PCRs in


parallel, with two different internal forward primers (LP13 and LP14), giving
full-length products that were larger (567 bp) or smaller (301 bp) than the 448 bp
ECO product, respectively. The spliced and unspliced products were also
matched for G+C content (Fig. II-5A), so a direct comparison would be reliable.
Moreover, these amplicons do not overlap the 5 GC-rich segment of Math5 that
is refractory to RNA and DNA polymerase processivity. Only the full-length
(unspliced) Math5 products were detected in these experiments by ethidium
bromide staining of agarose gels (Fig. II-5B). To consider this point more
rigorously, and detect extremely rare spliced Math5 mRNAs, we performed
identical competitive RT-PCRs with a common 6-carboxyfluorescein (6-FAM)
end-labeled reverse primer, and determined the molar ratio of spliced and
unspliced products by fluorescence capillary electrophoresis (Fig. II-5C). We
estimate that the ECO isoform represents less than 1.0% of Math5 transcripts in
the E14.5 eye. Math5 splicing thus does not occur at significant levels in the
developing mouse retina.
Direct test of Math5 splicing by nuclease protection (RPA)
To independently assess Math5 mRNA splicing in the retina, we
performed RNase protection assays (Melton et al., 1984). The nuclease
protection method was developed in the 1970s to demonstrate the existence of
mRNA splicing (Berget et al., 1978; Berk and Sharp, 1977; Eisenstein, 2005).
Unlike PCR, nuclease protection assays do not depend on an exponential
amplification process, which is highly sensitive to template secondary structure.

47

To evaluate the ratio of spliced and unspliced Math5 transcripts, we


hybridized total eye RNA from E14.5 embryos, in parallel, with a molar excess of
two 32P-labeled antisense RNAs (Fig. II-6A). These cRNAs were prepared by in
vitro transcription of two cDNA clones derived from unspliced 301 bp (A) and
567 bp (B) competitive RT-PCR products (Fig. II-5B). After hybridization and
RNase digestion, surviving probe RNA molecules were resolved by
polyacrylamide gel electrophoresis (Fig. II-6B). Probes A and B were protected
by full-length Math5 mRNA in the embryonic eye, giving 301 nt and 567 nt
digestion products. No hybridizing fragments were detected at the size predicted
for ECO mRNA (212 nt). The absence of smaller protected fragments in this
sensitive assay further indicates that the Math5 coding segment is not
significantly spliced in the embryonic eye.
Coding potential of lacunar Math5 RNAs
In addition to the ECO product, which lacks the entire coding region,
multiple mRNAs were proposed to originate from Math5 primary transcripts via
alternative splicing (Kanadia and Cepko, 2010). In some cases, these contain
partial open reading frames and were predicted to encode shorter Math5
isoforms. To test this hypothesis, the authors used commercial Math5 peptide
antisera to probe extracts from cells transfected with various splice products (cf.
Fig. 1H). We independently tested the reactivity of Math5 antibodies to mouse
and human proteins expressed at high levels in transfected NIH 3T3 cells, by
Western blotting, and to retinal sections from wild-type and Math5 mutant
embryos (Fig. II-S2), following standard precepts (Rhodes and Trimmer, 2006;

48

Saper and Sawchenko, 2003). We were unable to detect mouse Math5


polypeptide with any of these reagents, including the Abcam antisera (ab13536)
used by Kanadia and Cepko (2010).
Spliced Math5 transcripts in the cerebellum
During our previous characterization of Math5, we noted expression in the
developing hindbrain and cerebellum, including a single hybridizing Math5 mRNA
detected by Northern analysis (Saul et al., 2008). This 1.7 kb mRNA was
consistent with the size of embryonic retinal transcripts (Fig. II-2A). However,
three out of six Math5 clones derived from adult mouse brain RNA in the NCBI
database are apparently spliced, cerebellar (Cb) ESTs BY705389 and
AV030226, and cDNA AK005214 (Fig. II-S1A). These Cb isoforms are missing
199 nucleotides, and consequently are predicted to encode a truncated
polypeptide in which the 22 terminal amino acids of Math5
(VDPEPYGQRLFGFQPEPFPMAS) are replaced by 2 residues (VS). Although
the C-terminal amino acids are moderately conserved among amniotes (Fig. IIS3C), the donor splice sites are not.
To evaluate Math5 splicing in the cerebellum, we performed binary
(2-primer) and competitive RT-PCR experiments with total adult cerebellar RNA
as template (Fig. II-S3). In the binary PCR, the primers flanked the putative
199 bp intron, giving 567 bp (unspliced) or 368 bp (spliced) products (Fig. II-S3D).
In the triplex PCR, the two forward primers were located inside the intron and
spanning the exon junction (to amplify unspliced and spliced products
respectively), and the common reverse primer was end-labeled (Fig. II-S3E,F).

49

The reactions involved the terminal portion of the Math5 coding region and
3 UTR, and the products were similar in size (301 vs. 228 bp) and G+C content
(47.8 vs. 52.2 %). In contrast to the embryonic retina, we observed a moderate
level of alternative mRNA splicing in the adult cerebellum, involving 11 2 % of
Math5 transcripts. The major (1.7 kb) and minor (1.5 kb) cerebellar splice forms
were not previously resolved in Northern blots (Saul et al., 2008), presumably
because of the difference in abundance, and the effect of polyA tail heterogeneity
(with an expected mean length of 250 adenosines, (Wahle, 1995). This shorter
isoform was not detected in the embryonic retina (Fig. II-S3D).

Discussion
We have critically defined the transcriptional anatomy of the Math5 gene,
and characterized alternatively spliced mRNAs. In contrast to the adult
cerebellum, Math5 mRNA is not significantly spliced in the developing retina.
This conclusion is supported by six independent lines of evidence: Northern
analysis; RT-PCR analysis of natural and IVT-derived RNAs in the presence of
graded betaine concentrations; triplex RT-PCR analysis; EST data; and
ribonuclease protection assays. Our findings differ sharply from the recent report
of Kanadia and Cepko (2010). Three major factors contribute to the technical
artifacts observed by these authors: [1] intense secondary structure in the >85%
GC-rich segment of Math5 RNA and cDNA, which blocks the progression of
polymerase enzymes, creating a powerful negative selection; [2] RT template
switching in vitro; and [3] the existence of a vanishingly small population of

50

aberrantly spliced Math5 mRNAs (Fig. II-7A). In view of these results, further
investigation of Ngn3 splicing may be warranted (Fig. II-S4).
The GC-rich coding segment of Math5 (Fig. II-1B) evidently forms a
Gordian knot of secondary structure (Fig. II-7B,C), so dense that it favors the
amplification of minor cDNA products, representing less than 1% of Math5
molecules. G+C sequence bias is a well known problem in cDNA profiling studies
(Blackshaw et al., 2004; Margulies et al., 2001). The folded hairpin structure of
Math5 mRNA is relaxed in the presence of betaine. In vivo, local melting is
presumably catalyzed by DNA- and RNA-binding proteins, allowing Math5
replication, transcription and translation. However, the tight RNA secondary
structure may have consequences for Math5 protein expression. For example,
translation may require specific mRNA unwinding activity, creating another
potential mode of post-transcriptional regulation (Gray and Hentze, 1994).
Indeed, mRNA hairpins are known to impede ribosome elongation (Baim et al.,
1985) and G+C content is inversely correlated with translation efficiency
(Kenneson et al., 2001). If translation of the GC-rich Math5 mRNA were
hypersensitive to ribosome functional status, this may contribute to the disruption
of RGC development in Bst/+ mice, which have a mutation in the Rpl24
riboprotein gene and severe optic nerve hypoplasia (Oliver et al., 2004).
On the basis of these results, we believe that the most likely explanation
for the plethora of deleted Math5 cDNAs (Fig. II-4) is RNA template-switching
during the reverse transcriptase reaction, at points of sequence micro-homology
(Fig. II-7A, Table II-S3) (Brincat et al., 2002). Indeed, RT polymerases are

51

required to switch templates during normal retroviral replication, as part of the


first and second transfer steps (Telesnitsky and Goff, 1997). Aberrant switching
in vivo can generate intramolecular deletions, and the frequency is positively
correlated with the amount of RT pausing (Wu et al., 1995) and RNase H activity
(Brincat et al., 2002). In practice, template switching and related phenomena are
well known hazards in PCR-based expression studies, and have been
collectively termed RT-facts (Cocquet et al., 2006; Derjaguin and Churaev,
1973; Mader et al., 2001; Roy and Irimia, 2008; Zaphiropoulos, 2002).
The process of eukaryotic splicing produces a variety of functional and
nonproductive mRNAs during normal gene expression. While alternative splicing
greatly extends the genetic repertoire (Brett et al., 2002), particularly in the
nervous system (Li et al., 2007), a significant fraction of Pol-II transcripts are
mis-spliced, such that no protein or stable RNA species is synthesized, similar to
the ECO isoform. Frequent errors include exon skipping, intron retention, and
activation of cryptic splice sites. The resulting aberrant RNAs may outnumber
correctly spliced mRNAs among initial spliceosomal products (Jaillon et al., 2008;
Mitrovich and Anderson, 2000). For protein-coding genes with multiple exons, the
majority of aberrant RNAs contain a premature truncation codon (PTC) and are
degraded through the nonsense-mediated decay (NMD) pathway (Baker and
Parker, 2004). This is not generally possible for single-exon genes, which require
distinct quality control mechanisms to eliminate defective mRNAs (Maquat and
Li, 2001). The intronless class represents 5-15% of mammalian genes (Gentles
and Karlin, 1999; Sakharkar et al., 2004) and includes histones, GPCRs and

52

many Zn finger, HMG, and bHLH domain transcription factors.


The process of splice site recognition is also far more complicated than
the local pairing of 5 and 3 consensus sequences. It requires the holo definition
of exon or intron elements in context, with integration of multiple splice enhancer
and silencer effects (Berget, 1995; Fox-Walsh et al., 2005; Hertel, 2008; Wang
and Burge, 2008). In this way, intronless genes may have selectively acquired
sequence features that resist mRNA splicing (Fedorov et al., 2001; Irimia and
Roy, 2008; Jeffares et al., 2008). Detailed sequence comparisons of intronless
vs. intron-containing human genes has revealed differences in oligonucleotide
frequencies and context-dependent codon biases (Fedorov et al., 2001). The
most striking characteristic of intronless genes in this analysis was the overrepresentation of GC-rich 4- to 6-mers, after correcting for base composition. The
Math5 cDNA matches this pattern extremely well (not shown), exhibiting
sequence features that are characteristic of intronless genes. Moreover, the
GGG triplet, which binds U1 snRNP as an intronic splice enhancer (Engelbrecht
et al., 1992; McCullough and Berget, 2000), is depleted within the Math5 coding
region, despite the high G+C content. These global compositional features are
not considered by the Spliceport algorithm that was used by Kanadia and Cepko
to predict Math5 splice sites. This web-based tool performs statistical analysis of
k-mers in a 160 nt window surrounding putative donor and acceptor sites, based
on human genome search data (Dogan et al., 2007). The analysis predicted the
alternative Cb splice acceptor, which is utilized at low frequency in the adult
cerebellum (FGA score = 1.33); however, the Cb donor site was not identified

53

and statistical support for donor sites in the Math5 transcript was relatively low
(max FGA score = 0.26). Indeed, the mouse genome contains many more weak,
potential splice sites than are actually utilized in vivo.
Among the numerous Math5 species reported by Kanadia and Cepko,
only one PCR product, termed ECO, is compatible with mRNA splicing. On the
basis of our results, we believe this solitary cDNA is derived from an aberrantly
spliced transcript, which has escaped normal quality control. First, the RNA
encodes no protein and has no demonstrated function. In other contexts, long
ncRNAs such as Xist and Air, have been shown to have regulatory roles (Mercer
et al., 2009), and a small number of bifunctional mRNAs have alternate coding
and noncoding isoforms (Chooniedass-Kothari et al., 2004). Second, the ECO
isoform is very rare, representing less than 1% of Math5 mRNA, and is thus
unlikely to have a significant role in regulating Math5 function or modulating
retinal cell fate determination.
An intriguing result from our study is the discovery that 11% of mature
Math5 transcripts in the adult cerebellum are bona fide spliced mRNAs. These
are predicted to encode a shorter Math5 protein, which lacks 20 amino acids
from the C-terminus and may exhibit unique molecular properties (Fig. II-S3).
However, its function is not known, and Math5 mutants have no overt cerebellar
phenotype (Saul et al., 2008).
Despite the intriguing hypothesis advanced by Kanadia and Cepko, our
results show splicing of Math5 mRNA into noncoding isoforms does not occur in
the developing retina at levels greater than 1% of transcripts. Further studies are

54

needed to determine the exact mechanism of Math5 action, how progenitors are
transformed into neurons, and how noncoding RNAs, including microRNAs, may
regulate Math5 expression, RGC development, and the diversification of ganglion
cell subtypes.

Acknowledgements
The authors are grateful to John Moran and David Turner for helpful
suggestions; to Dellaney Rudolph, Tien Le, Susan Tarl and the UM sequencing
core for technical support; and to John Moran, David Turner, Miriam Meisler,
Doug Engel, Chris Chou, and Terri Grodzicker for careful reading of the
manuscript. The research was funded by NIH R01 grants to TG (EY14259) and
NLB (EY13612) and The Glaucoma Foundation (TG). LP was supported by NIH
T32 grants to the University of Michigan Medical Scientist (GM07863) and Vision
Research (EY13934) Training Programs.

55

Figure II-1. Anatomy of the Math5 transcription unit. (A) Gene map showing the
major 1489 nt mRNA species; coding region (red box) and UTRs; direct repeats
(DR); major polyA signal (pA) and internal A-rich segment (A14); cerebellarspecific intron (Cb); and PCR primers used in this study (dark red). LP15 spans
the Cb intron junction. (B) Plot showing elevated GC content (red) across the
Math5 coding region, compared to the average value (49.98%) for the mouse
transcriptome (green) (Stolting et al., 2009). The 150 nt segment with >85% GC
and the 536 nt fold encompassing the coding region are indicated (brackets). (C)
Concentration of polymerase-refractory YGC trinucleotides in the proximal coding
region (both strands). (D) Magnified view of the Math5 promoter showing the
TATAA box, transcription start site (TSS) and 5 termini of cDNA clones (Brown
et al., 2002; Brown et al., 1998). (E) Sequence of UTR direct repeats.

56

57

Figure II-2. Math5 messenger RNAs. (A) Northern blot probed with 1.2 kb Math5
(JN4C) and 1.1 kb -actin cDNAs. Two Math5 mRNAs are visible (left
arrowheads), but no hybridizing RNA species is present in the 0.8-1.0 kb size
range. The RNA size ladder cross-hybridized to vector DNA in the plasmid
probes. (B) Map of the 3 UTR and flanking genomic DNA (6 kb), showing eight
potential polyA signals ATTAAA (blue) and AATAAA (green); the internal A14
priming site in the UTR (pA); interspersed repeats (gray); and the nested
3 RACE primers (dark red) for pA1 and pA6 sites, which have the most favorable
sequence context. Clones JN2 and BC092234 terminate at pA1, whereas
cDNAs JN1, JN4 and JN6 terminate at pA (Brown et al., 2002; Brown et al.,
1998). pA2 marks an A-rich genomic site captured in the pA6 assay. (C)
polyADQ scores for all potential pA sites, calculated using human genome
parameters (Tabaska and Zhang, 1999). Only pA1 and pA6 have scores above
threshold. (D) Embryonic eye RT-PCRs with 260 bp and 365 bp 3 RACE
products (arrowheads) showing utilization of pA1 and pA6 sites. The 900 bp
product was primed from pA2 (open arrowhead). m, marker (1 kb-plus ladder);
RT, reverse transcriptase. (E) Sequence of pA1 RACE products originating from
the 1.7 kb Math5 mRNA. (F) Sequence of pA6 RACE products originating from
the 4.4 kb Math5 mRNA.

58

59

Figure II-3. Math5 embryonic eye RT-PCRs with increasing amounts of betaine. (A)
Agarose gel showing cDNA products amplified from DNase-treated E14.5 eye RNA
with UTR primers LP8 and LP4 in the presence of 0X, 1X, 2X and 3X MasterampTM.
When the betaine concentration was increased, only the full-length 1087bp Math5
cDNA product was visible; the 448bp ECO product was absent. No amplimers were
observed in the absence (-) of RNA template or RT enzyme. The identity of all PCR
products was verified by sequencing. (B) Similar PCR with a mouse genomic DNA
template, showing amplification of the identical full-length 1087bp product. (C,D)
Parallel PCRs were performed using internal primers LP6 and LP7. A single 486bp
Math5 product was amplified from cDNA or gDNA in 2-3XMasterampTM.

60

Figure II-4. RT-PCRs of Math5 RNA transcribed in vitro. (A) Diagram and
agarose gel showing linearized pJN4C and Math5 sense RNA generated by T3
polymerase and treated with DNaseI. (B) cDNA products amplified by RT-PCR
from IVT-derived RNA with UTR primers LP8 and LP4. Only the full-length
1087 bp Math5 cDNA product was amplified in the presence of 3X Masteramp
(MA, indicated above brackets). In the absence of betaine, a variety of weak
products were observed, with a heterogeneous deletion profile, reflecting a low
level of RT template switching. This background could be increased by using
suboptimal PCR conditions or omitting the mouse liver RNA carrier. IVT, in vitro
transcribed Math5 RNA (10 ng); ML, mouse liver RNA (3 g). (C) Similar
RT-PCRs performed using internal primers LP6 and LP7. Only the expected
486 bp cDNA was amplified in 3X MA, while spurious products were amplified at
lower MA concentrations. The right three panels in B and C represent adjacent
lanes in the same gels, displayed separately for clarity. (D) Alignment of lacunar
cDNAs generated from IVT or E14.5 eye RNA templates. The deletion profile is
comparable to the distribution reported by Kanadia and Cepko (2010, cf. Suppl.
Table 1 and Figure 1), using the same primer pairs with no precautions for GC
secondary structure. The sequence of breakpoints is given in Table S3, with
microhomology at the inferred sites of RT template switching.

61

62

Figure II-5. Triplex competitive RT-PCR assay to evaluate trace levels of Math5
splicing in the embryonic retina. (A) Diagram showing PCR strategy. The length
and % G+C of competing amplicons are comparable. (B) Agarose gel stained
with ethidium bromide, showing only the unspliced Math5 cDNA product in each
assay. (C). Capillary electrophoresis profiles showing triplex competitive
RT-PCR products (top panels) and the ECO product amplified with duplex UTR
primers in the presence of 1X MA (bottom panel). The common antisense primer
(LP4) was end-labeled with 6-FAM. From the peak areas measured in replicate
experiments and mixing controls, we estimate that the ECO product represents
0.4 to 1.0 percent of Math5 mRNA in the embryonic retina, which is near the
detection limit of this assay.

63

64

Figure II-6. Ribonuclease protection assays. (A) Diagram showing RPA strategy, with
Math5 cDNA, two different antisense cRNA probes, protected fragments expected for
FL (full length, unspliced) and ECO (spliced) transcripts, and positive control RNAs
generated by sense IVT reactions. (B) Autoradiogram, showing undigested probes A
and B (366 nt and 632 nt) and exclusively unspliced fragments protected by E14.5 eye
RNA (567 nt and 301 nt). No fragment corresponding to the presumptive ECO transcript
(212 nt) was protected by eye RNA using either cRNA probe, although a doublet of this
size was protected by the ECO IVT positive control. Background fragments observed
with probe B (arrowheads) are caused by intrinsic sensitivity of the cRNA-mRNA duplex
to RNase cleavage at particular sites and were also present in the full length IVT
positive control. The probe (no RNase) and IVT controls were diluted 20- and 10-fold
respectively, compared to the E14.5 eye RNA hybridization lanes.

65

66

Figure II-7. Model explaining the observed results. (A) Diagram showing the
likely origin of heterogeneous deleted Math5 cDNAs, through combined effects of
RT template switching, trace levels of aberrantly spliced ECO mRNA, and
powerful PCR selection favoring deletion of GC-rich coding sequences. (B)
Secondary structure predicted for the major 1489nt Math5 mRNA. This M-fold
circle diagram, generated by free energy (G) minimization, is magnified in Figure
S5. Red, blue and green arc lines indicate G-C, A-U and A-G base pairs. The
coding region, DRs and presumptive ECO splice sites are labeled. The 150nt
segment described in the text with >85% G+C, and the segment expanded in
panel C are marked. (C) Stem-loop diagram showing the 536nt fold that
encompasses the Math5 CDS with lowest free energy (G=-258 kcal/mol) and
Tm 82C. The major structural features in panels B and C are labeled alike. (D)
Junctional sequences for the ECO product with presumptive splice sites,
compared to the U1 consensus.

67

Figure II-S1. Math5 ESTs in the public domain. (A) Diagram modified from the
UCSC mouse genome browser (mm9 assembly, chr10 : 62,562,000 - 62,564,300)
showing 56 Math5 ESTs and 2 Genbank cDNAs (BC092234, AK005214), giving
a total n = 58, with 52 derived from the embryonic retina. Forty-three of these
retinal cDNAs cross the presumptive ECO junctions at the 5 or 3 side, and are
thus informative for splicing (83 %). Yet none originated from spliced mRNA. Of
the remaining six, from adult brain RNA (red), two cerebellar ESTs and one
cDNA were spliced at the Cb intron (yellow shading, see Figure S3). Nine 3
ESTs out of 21 terminate at pA1; the remaining 12 were primed from pA. (B)
Comparable region of the human genome (hg19 assembly,
chr10 : 69,992,300 - 69,990,000) showing one full-length Genbank cDNA and 7
unspliced ESTs.

68

69

Figure II-S2. Evaluation of Math5 antibodies. (A) Diagram of the mouse Math5
protein, showing the antigenic index (Jameson and Wolf, 1988) and positions of
immunogens used by various sources to prepare antibodies, as follows: a,b
internal and C-terminal peptides (Glaser lab); c, ab13536 (Abcam); d, AB5694
(Chemicon); e, EB07972 (Everest); f, 1A5 (multiple vendors). The immunogens
for D01P (Abnova) and MAb 1A5 were full-length or partial recombinant human
proteins (gray); all others were based on the mouse polypeptide (blue). No
immunogen was specified for ab78046 (Abcam). (B) Immunoblots of NIH3T3
cells co-transfected in parallel with pUS2-EGFP and pCS2 expression plasmids
for full-length mouse or human Math5 proteins six N-terminal Myc epitope tags,
or empty pCS2 vector. Five identical blots were probed using antibodies with
stated reactivity to mouse (ab13536, ab78046) or human (D01P) Math5; Myc or
GFP. The predicted mass for native and 6xMyc mouse Math5 proteins is 16.9
and 27.0 kDa, respectively. Antibody D01P detected the human polypeptides, but
not mouse. No other reagent tested was effective, including ab13536 (Abcam)
(Kanadia and Cepko, 2010), even when the Math5 proteins were massively
overexpressed. (C) Retinal sections from E15.5 embryos immunostained with
ab13536 sera. The immunofluorescence pattern was identical between wild-type
and Math5 -/- eyes and is thus nonspecific (Rhodes and Trimmer, 2006; Saper
and Sawchenko, 2003). This pattern, which includes lens and RPE nuclei, does
not fit the apical distribution of Math5 mRNA in the neuroblastic retina. The in situ
hybridization pattern of a Math5 cRNA probe spanning the 3 UTR and CDS
matches our previous reports (Brown et al., 1998; Hufnagel et al., 2010) and both
panels provided by Kanadia and Cepko (cf. Figure 1j and 1j).

70

71

Figure II-S3. Math5 splicing in the cerebellum. (A) Diagram of alternative Cb


intron, with PCR primers and products. (B) Sequence of Cb splice junction,
corresponding to nucleotides 3524 and 3724 in Genbank acc. AF418923. The
acceptor site coincides with the ECO junction (Figure 7D). (C) Spliced cerebellar
mRNA encodes a truncated Math5 protein, with 20 fewer amino acids at the Cterminus. The deleted peptide has a similar sequence among amniotes, but the
splice junction is not obviously conserved. (D) Agarose gel showing spliced
(368 bp) and unspliced (567 bp) RT-PCR products from the adult cerebellar RNA,
but not from E14.5 retina. (E) Triplex competitive RT-PCR showing spliced
(228 bp) and unspliced (301 bp) products co-amplified from cerebellar cDNA
(right lanes). In the duplex control with primers LP15 and LP4, only the Cb form
(spliced) was amplified (left lanes). (F) Capillary electrophoresis profiles showing
the ratio of spliced (Cb) and unspliced (FL) transcripts in the triplex PCR (top),
with Cb duplex product as a control (bottom). The common antisense primer LP4
was labeled with 6-FAM. Approximately 11 2 percent of Math5 mRNAs are
spliced at the Cb site in the adult cerebellum.

72

73

Figure II-S4. Splicing patterns in the mouse Atonal-related bHLH genes. (A)
Phylogram of mouse Ato proteins, based on maximum parsimony analysis of the
bHLH domain across many taxa (Blackburn et al., 2008; Brown et al., 2002). (B)
Exon-intron organization of bHLH genes based on a survey of ESTs in the NCBI
database (Brett et al., 2002; Harrington et al., 2004). The eight mouse Ato
homologs either have unitary exon structures, or a single intron located in the
5'UTR. The Achaete-Scute homolog Mash1 (Ascl1) has a single intron in the
3'UTR. There is no obvious correlation between splicing patterns and locations in
the mouse genome. MMU, mouse chromo ESTs, number of expressed sequence
tags supporting the gene structure; *has minor alternative spliced product (Cb);
**has overlapping intergenic and antisense RNAs. The intron of one spliced
antisense EST (CF104925) for Ngn3 (Neurog3) overlaps the 5'UTR and coding
sequence of the sense strand. This antisense RNA is predicted to co-amplify in the
RT-PCR and may be mistaken for non-coding sense products.

74

Figure II-S5. Secondary structure for Math5 mRNA. This circle plot was
generated by free energy minimization of the 1489nt mRNA, and is enlarged
from Fig. II-7. Red, blue and green arc lines indicate G C, A U and A G base
pairs. The coding region, DRs and presumptive ECO splice sites are labeled.
The 150nt segment with >85% G+C, and 536nt segment spanning the CDS
are marked. The CDS contains a high density of G-C base pairs (red arcs),
which are deleted in rare, mis spliced RNAs.

75

76

S
AS
AS
AS
S
S
AS
S
AS
AS
AS
S
S
S
S
S
S

LP1
LP2
LP3
LP4
LP5
LP6
LP7
LP8
LP9
AP
UAP
LP10
LP11
LP12
LP13
LP14
LP15

TCTACTGCAAGCTGTCCAAACGCTC
AACATACAGGCTGTGTTGGTAGCTG
GGTAGCTGCTCAGAACATAAACAAGTCACAT
GTTTCTCCACCTCCTGAATGACGCT
GCCTCCCTATCTCCACTTCTCTTGTGT
GTGGATGAAGTCGGCCTGCAAA
TTTCTCCCCTAAGACCCAAATGGC
TCTCAGGCTTTCCCAGAGAACTGGA
TTTGCAGGCCGACTTCATCCAC
GGCCACGCGTCGACTAGTACTTTTTTTTTTTTTTTTT
GGCCACGCGTCGACTAGTAC
TCCCTATTGGGCGAAGTTGT
AGGGTGAAGTGCTTGCTGGT
GTTACAGGGCCTGCGAAATG
AAGCTGTCCAAGTACGAGACACTGC
CCTTTTCTGCTTAATTTCCTTCCCG
GGGTGCTAGGCTCCAG|GTTTC

Sequence (5 3)
RT-PCR
RT-PCR
RT-PCR
RT-PCR, triplex PCR
RT-PCR, 3RACE (pA1)
RT-PCR
RT-PCR
RT-PCR, triplex PCR
RT-PCR
3RACE
3RACE
3RACE (pA1)
3RACE (pA6)
3RACE (pA6)
RT-PCR, triplex PCR
triplex PCR
triplex PCR (Cb)*

Expt
1
2
3
4
5
6
7
1
2

Alternate names
Sup. Fig. 2
Sup. Table 2

Ori, orientation; S, sense; AS, antisense; AP, adapter primer; UAP, universal amplification primer
* primer sequence spans Cb intron junction
end-labeled with 6-carboxyfluorescein (6-FAM)
used by Kanadia and Cepko (2010)

Ori

Name

Table II-S1. Oligonucleotide primers in this study

77
LP4*

LP4*

LP4*

UAP
UAP

UAP
UAP

LP4*

LP4

LP7

33

33
33

33

15
20

15
20

33

35

35 or 40

35 or 40

# cycles

unspliced
Cb splice

unspliced
ECO

unspliced
ECO

initial RACE
nested RACE

initial RACE
nested RACE

Cb splice

unspliced
Cb splice

Math5 full length


ECO

Product

301
228

301
448

567
448

>379
>342

>591
>236

228

567
368

486

1087
448

Size

Suppl 3e

6b, 6c

6b, 6c

2d

2d

Suppl 3e

Suppl 3d

3c, 3d, 4c

3a, 3b, 4b

Figure

Products include a variable polyA tract


Triplex PCRs untilized 3 primers, with 2 FOR primers at 0.1 M each and 1 REV primer at 0.2 M.

All PCRs had an initial denaturation step (94C x 3 min); followed by # cycles of 94C x 30 sec
denaturation, 57C x 45 sec annealing, and 72C x 60-70 sec extension; with a final extension
step (72C x 7 min)

* End-labeled with 6-FAM (carboxyfluorescein)

Notes:

LP14
LP15

LP11
LP12

3RACE pA6

Triplex RT-PCR

LP5
LP10

3RACE pA1

LP8
LP14

LP15

RT-PCR

Triplex RT-PCR

LP13

RT-PCR

LP8
LP13

LP6

RT-PCR, gPCR

Triplex RT-PCR

LP8

RT-PCR, gPCR

LP4

Primers
FOR
REV

Experiment

Table II-S2. PCR conditions used in this study

78

5 sequence

...
...
...
...
...

ACTGACTGCAC|GTGAGTCGCTCCGTC
CCCTCCGGCGGGAGCT|CGCGGCGCGC
TCGCGGCGCGCCCCCGT|GCGCGGGCG
TCGCGGCGCGCCCCCG|TGCGCGGGCG
GGCCCTCCGGCGGGAGCTCGC|GGCGC
TM

[639]
[71]
[85]
[134]
[122]

del

TGTTTTGTAG|GTTTCATGAGTGGACATC
|CGCGGCGCGCAGGCGTCTGGCGGCCAAC
|GCGCGAGCGGCGCCGCATGCAGGGGCTG
|TGCGCAGGGTGGTGCCGCAGTGGGGCCA
GG|GGCTGAACACGGCGTTCGACCGGCTG

3 sequence

... AAGTCGGCCTGCAAACCCCAC|GGCCC
... ATGAAGTCGGCCTGCAA|ACCCCACG

...
...
...
...
...
...
...
...
...
...

CGTGGATGAAGTCGGCCTG|CAAACC
GAAAGGCTTTCTAT|CCCCGACCCCC
CTCCGGCGGGAGCTC|GCGGCGCGCC
CTCCGGCGGGAGCTC|GCGGCGCGCC
CTCCGGCGGGAGCTC|GCGGCGCGCC
CGCGCCCCCGTGCGCGG|GCGCAGCC
CTCGCGGCGCGCCCCCG|TGCGCGGGC
CCCCACGGCCCTCCGGCGG|GAGCTCG
CCCTCCGGCGGGAGCT|CGCGGCGCGC
TGCAAACCCCACGGCCCT|CCGGCGGG

[746]
[605]
[106]
[106]
[70]
[85]
[103]
[110]
[71]
[155]

[70]
[235]

|TAATCCTAGCGTCATTCAGGAGGTGGA
|CCCCTACCTCCCTTTCCCGGGTGCTAG
|GCGGCGCCGCATGCAGGGGCTGAACAC
|GCGGCGCCGCATGCAGGGGCTGAACAC
|CGCGGCGCGCAGGCGTCTGGCGGCCAA
GCGCGA|GCGGCGCCGCATGCAGGGGCT
|TGCAGGGGCTGAACACGGCGTTCGACC
GCGC|GAGCGGCGCCGCATGCAGGGGCT
|CGCGGCGCGCAGGCGTCTGGCGGCCAA
|CCGGCTGCGCAGGGTGGTGCCGCAGTG

...
...
...
...
...
...
...
...
...
...

|GGCCCGGGCGGCTGGAGAGCGCGGCGCG ...
CTGCAG|ATGGCGCTCAGCTACATCATCG ...

...
...
...
...
...

* Product is likely due to mis-priming of LP6.

Spurious RT-PCR products are numbered to match Figure 5. The highlighted text shows
areas of micro-homology, which can promote RT template switching. The size of the
deletion [bp] and breakpoints (|) are indicated.

3
4
5
6
7
8
9
10
11
12

from RT-PCR of IVT RNA ( 2X MasterAmpTM)

1
2*

from E14.5 retina RT-PCR ( 2X MasterAmp , 40 cycles)

ECO
InC
E12P20
32
9

from Kanadia and Cepko (2010)

No.

Table II-S3. DNA sequence flanking deletions in RT-PCR products

CHAPTER III

MATH5 DEFINES THE GANGLION CELL COMPETENCE STATE IN A


SUBPOPULATION OF RETINAL PROGENITOR CELLS EXITING THE CELL
CYCLE

Abstract
The basic helix-loop-helix (bHLH) transcription factor Math5 (Atoh7) is
transiently expressed during early retinal histogenesis and is necessary for
retinal ganglion cell (RGC) development. Using nucleoside pulse-chase
experiments and clonal analysis, we determined that progenitor cells activate
Math5 during or after the terminal division, with progressively later onset as
histogenesis proceeds. We have traced the lineage of Math5+ cells using mouse
BAC transgenes that express Cre recombinase under strict regulatory control.
Quantitative analysis showed that Math5+ progenitors express equivalent levels
of Math5 and contribute to every major cell type in the adult retina, but are
heavily skewed toward early fates. The Math5>Cre transgene labels 3% of cells
in adult retina, including 55% of RGCs. Only 11% of Math5+ progenitors develop
into RGCs; the majority become photoreceptors. The fate bias of the Math5
cohort, inferred from the ratio of cone and rod births, changes over time, in
parallel with the remaining neurogenic population. Comparable results were
obtained using Math5 mutant mice, except that ganglion cells were essentially

79

absent, and late fates were overrepresented within the lineage. We identified
Math5-independent RGC precursors in the earliest-born (embryonic day 11)
retinal cohort, but these precursors require Math5-expressing cells for
differentiation. Math5 thus acts permissively to establish RGC competence
within a subset of progenitors, but is not sufficient for fate specification. It does
not autonomously promote or suppress the determination of non-RGC fates.
These data are consistent with progressive and temporal restriction models for
retinal neurogenesis, in which environmental factors influence the final histotypic
choice.

Introduction
The seven major cell types in the vertebrate retina (rod and cone
photoreceptors; amacrine, horizontal and bipolar interneurons; Mller glia; and
ganglion cells) develop from a common pool of progenitors (Turner and Cepko,
1987; Turner et al., 1990) that are established when the optic vesicles invaginate
to form bilayered optic cups (Goldowitz et al., 1996). The inner layer of each
optic cup consists of proliferative retinal progenitor cells (RPCs), which are
arranged as a pseudostratified epithelium. These RPCs begin to permanently
exit mitosis and differentiate around embryonic day 11 (E11) in the mouse.
Retinal neurons and glia are fully formed by postnatal day 21 (P21) and are
arranged in a highly ordered tri-laminated structure (Rodieck, 1998). The outer
nuclear layer (ONL) consists of photoreceptors while the inner nuclear (INL) and
ganglion cell (GCL) layers are populated by interneurons, glia and ganglion cells.

80

The mechanism of cell fate determination how these diverse cell types are
generated from an initially homogeneous progenitor population remains poorly
understood.
Birthdating experiments, in which [3H]-thymidine was used to mark the
terminal S phase of progenitor cells, have established a characteristic order for
the emergence of different retinal cell types during histogenesis (Carter-Dawson
and LaVail, 1979; Rapaport et al., 2004; Sidman, 1961; Young, 1985a). In all
vertebrate species examined, retinal ganglion cells are the first-born neurons
(Altshuler et al., 1991). In mammals, these are followed by horizontal cells,
cones, amacrines, rods, bipolar cells and Mller glia, in descending birth order.
There is considerable overlap in the distribution of birthdates among cell types,
particularly for rod photoreceptors, which are born over an extended period (E13P7 in mice) and are most abundant. Moreover, as a subclass, displaced
amacrines, located in the mammalian GCL, are born earlier than amacrines in
the INL (LaVail et al., 1991; Reese and Colello, 1992).
Lineage tracing experiments in rodents and frogs show that individual
retinal progenitors are multipotent, giving rise to clones with heterogeneous cell
type composition and size, and that the histogenic potential of the progenitor pool
is gradually restricted over time (Holt et al., 1988; Turner and Cepko, 1987;
Turner et al., 1990; Wetts and Fraser, 1988; Wong and Rapaport, 2009). The
absence of a strict hierarchical relationship among cell types suggests that fate
determination in the retina is a stochastic process (Gomes et al., 2011; Livesey
and Cepko, 2001). The observation of discordant two-cell clones in rodent

81

lineage marking studies indicates that at least some cell fate decisions occur
during or after the terminal division, and may be subject to environmental
influence (Turner and Cepko, 1987). Indeed, multiple extrinsic factors have been
shown to alter the ratio of retinal cell types generated from progenitor pools
(Altshuler et al., 1991; Ezzeddine et al., 1997; Fuhrmann et al., 1995; Yang,
2004; Young and Cepko, 2004).
Heterochronic mixing experiments, in which early and late retinal cells are
co-cultured in unequal ratios, have shown that progenitors have a limited
capacity to shift their fate forward or backward in sequence, and suggest that
competence is fundamentally a cell-intrinsic property (Belliveau and Cepko,
1999; Rapaport et al., 2001; Reh, 1992; Watanabe and Raff, 1990). Likewise,
single-cell dissociation studies have shown that the fates of retinal progenitors,
including post-mitotic cells, change over time and are intrinsically programmed
(Adler and Hatlee, 1989; Cayouette et al., 2003; Reh and Kljavin, 1989). Thus, it
is likely that cell-intrinsic factors, expressed by progenitors in a prescribed
temporal order, work in concert with extrinsic factors in the retinal
microenvironment to guide cell fate decisions and ensure proper ratios of each
cell type.
The basic helix-loop-helix (bHLH) transcription factor Math5 (Atoh7) was
identified on the basis of its homology to Drosophila Atonal (Brown et al., 1998),
which plays a critical role in the specification of R8 photoreceptors in the eye
imaginal disc (Frankfort and Mardon, 2002; Hsiung and Moses, 2002; Jarman,
2000; Sun et al., 2003). The mouse Math5 gene contains a single exon (Prasov

82

et al., 2010) and is specifically expressed by progenitor cells during retinal


histogenesis (Brown et al., 1998), similar to frog, chick, and zebrafish orthologs
(Kanekar et al., 1997; Liu et al., 2001; Masai, 2000). Math5 mutant mice lack
retinal ganglion cells (RGCs) and optic nerves (Brown et al., 2001; Wang et al.,
2001) and their circadian rhythms are not photoentrained (Brzezinski et al., 2005;
Wee et al., 2002). Retinal vascular development (Brzezinski et al., 2003) and
electrophysiology (Brzezinski et al., 2005) are also disrupted in these mice.
Finally, the relative abundance of other retinal cell types is altered, through a
combination of cell autonomous and non-autonomous effects (Brzezinski et al.,
2005; Le et al., 2006). RGC genesis similarly fails in ath5 mutant (lakritz)
zebrafish (Kay et al., 2001). In humans, ATOH7 mutations cause optic nerve
aplasia (Ghiasvand et al., 2011) and the ATOH7 locus is a major determinant of
normal variation in optic disc size, which reflects RGC number (Khor et al., 2011;
Macgregor et al., 2010; Ramdas et al., 2010).
Math5 is likely to trigger a regulatory cascade for RGC development.
Expression of the POU domain transcription factor Brn3b (Pou4f2) appears to be
controlled by Math5 in mice, similar to the orthologous circuit in chick and frog
(Hutcheson and Vetter, 2001; Liu et al., 2001; Schneider et al., 2001; Wang et
al., 2001). In turn, Brn3b and the homeodomain transcription factor Isl1 form two
regulatory nodes that are critical for RGC maturation (Erkman et al., 1996; Gan
et al., 1996; Mu et al., 2004; Mu et al., 2008; Pan et al., 2008).
How does Math5 regulate ganglion cell fate determination? In principle,
Math5 could act either as an instructive factor, irreversibly directing competent

83

progenitors to differentiate into RGCs, or as a permissive factor, establishing an


RGC competence state within a set of multipotent progenitors, only some of
which develop into RGCs (Wessells, 1977). The Cre-lox recombination system
provides a powerful tool to distinguish these mechanisms, by indelibly marking
descendant cells. In a previous lineage analysis, a Math5-Cre knock-in allele
was found to mark multiple retinal cell types, suggesting that Math5 acts
permissively (Feng et al., 2010; Yang et al., 2003).
In this report, we extend these findings using a Math5>Cre BAC transgene
in wild-type and Math5 mutant mice. This approach, coupled with birthdating
analysis, has allowed us to quantitatively assess the cell type distribution and
unique fate trajectory of the Math5 lineage over time. Our results show Math5 is
expressed at equivalent levels in a subset of progenitors that are capable of
forming all retinal cell types, with a frequency that decreases according to birth
order. Although heavily weighted toward early fates, only 11% of these cells
develop into RGCs and only 55% of RGCs descend from Math5+ progenitors. In
the absence of Math5 function, lineage-marked cells exhibit a similarly diverse
range of fates but do not differentiate as RGCs, suggesting Math5 has both
autonomous and non-autonomous roles in RGC development. Using cell cycle
markers and nucleoside pulse-chase analysis, we show Math5 expression is
confined to progenitors during or after the terminal division, and does not control
cell cycle exit. Finally, using retroviral clone analysis of explanted embryonic
retinas, we demonstrate that Math5+ cells frequently arise in pairs from
symmetric terminal divisions. Our results extend previous observations, but

84

compel different conclusions. We provide new insights into Math5 function,


ganglion cell development, and the mechanism of retinal fate determination.

Materials and Methods


Quantitative PCR
Eye tissue was collected from 8-12 CD-1 embryos or newborn mice at
time-points between E10.5 and P1.5 and homogenized in Trizol reagent
(Invitrogen, Carlsbad, CA). Total RNA was purified from pooled homogenates at
each time-point. cDNA was synthesized using d(N)6 primer and Superscript II
reverse transcriptase (Invitrogen). Quantitative PCR was performed on cDNA
using Math5 and Hprt primers (Brown et al., 2001) with the iCycler iQ system
(Bio-Rad, Hercules, CA). Seven measurements were made for each cDNA pool.
Math5 RNA levels (critical threshold cycles) were normalized to Hprt as
described (Livak and Schmittgen, 2001), and are reported relative to the mean
P1.5 value.
Math5>Cre BAC transgenic mice
We replaced the Math5 open reading frame on bacterial artificial
chromosome (BAC) clone RP23-328P3 with a 2.0 kb nlsCre-actin pA cassette
using a two-step recA-mediated recombination protocol in E. coli (Gong et al.,
2002; Heintz, 2001). To target the BAC, which contains 110 kb 5 and 103 kb 3
genomic DNA flanking the Math5 transcription unit (Prasov et al., 2010), we
constructed a plasmid vector with short 5 (A, 345 bp) and 3 (B, 378 bp)
homology arms flanking Cre-pA. These were amplified by PCR from UTR

85

sequences of the solitary Math5 exon (AF418923) and cloned into the SalI and
XhoI sites of pGSU-Cre (Cushman et al., 2000). The resulting A-Cre-B
cassette was inserted into the XhoI site of shuttle plasmid pLD53GFP10 as a
2.4 kb SacI-XhoI fragment and verified by DNA sequencing. Shuttle plasmid
pLD53GFP10 was derived from pLD53.SC1 by partial SpeI digestion and
insertion of a XhoI linker in place of the 3.5 kb EGFP fragment. We then targeted
RP23-328P3 with the Math5>Cre shuttle vector pLD53ACreB to obtain
ampicillin- and chloramphenicol-resistant cointegrates (Gong et al., 2002).
These were resolved by selection on TYE (tryptone-yeast extract) agarose plates
with chloramphenicol and 10% (w/v) sucrose. Two recombinant Math5>Cre BAC
clones were recovered and verified by PCR and pulsed-field gel electrophoresis
(PFGE) Southern analysis.
Purified circular DNA from Math5>Cre BAC clone RP23-328P3-D1-68 was
injected into fertilized (SJL/2 C57BL/6J) F2 oocytes by the UM Transgenic
Animal Core Facility. Nine transgenic founders were identified by Cre-specific
and BAC vector-insert junctional PCRs. Transgene copy number was
determined by Southern analysis, using an upstream Math5 genomic probe that
hybridizes equally well to 3.5 kb BAC and 6.5 kb mouse chromosomal EcoRI
fragments. Transgene integrity was evaluated by Southern analysis following
NotI digestion and PFGE. Transgenic offspring were genotyped using PCR
primers within the Cre-pA cassette.
Math5>Cre mice (line 872 or 360) were crossed to Z/AP (JAX stock
003919, (Lobe et al., 1999) and R26floxGFP (JAX stock 004077, (Mao et al.,

86

2001) reporter strains, which express membrane-tethered hPLAP (human


placental alkaline phosphatase) or cytoplasmic GFP (green fluorescent protein),
respectively, from ubiquitously active promoters, upon Cre-mediated excision of
floxed upstream stop signals. Tissues from informative double transgenic
progeny were collected from E11.5 to 15.5, on P0.5, and at 3-4 weeks of age.
To trace lineage in the absence of Math5 function, we crossed Z/AP; Math5 -/mice (Atoh7tm1Gla, Brown et al., 2001) to Math5>Cre (line 360); Math5 +/- mice
and compared the patterns of hPLAP staining in 3-4 week-old double transgenic
mutants and heterozygous controls.
Histology
Embryonic and adult eyes were fixed overnight in 4% paraformaldehyde
(PFA) at 4C, cryoprotected in phosphate-buffered saline (PBS) with 10 to 30%
sucrose, frozen in OCT compound (Tissue-Tek, Torrance, CA), and
cryosectioned at 5-10 m. For Brn3b (Pou4f2) and cyclin D1 epitopes, fixation
was 30 min at room temperature in 2% PFA. For immunodetection, cryosections
were blocked for 4 hrs at room temperature in PBTx (0.1 M NaPO4 pH 7.3 0.5%
Triton X-100) with 10% normal donkey serum (NDS) and 1% bovine serum
albumin (BSA). Sections were incubated overnight at 4C with primary antisera
or biotinylated PNA (peanut agglutinin) lectin diluted in PBTx with 3% NDS and
1% BSA. For fluorescence detection, sections were incubated for 2 hrs at room
temperature with appropriate secondary antibodies or streptavidin conjugates
(Jackson Immunoresearch, West Grove, PA). Nuclei were identified using 100
ng/mL 4',6-diamidino-2-phenylindole (DAPI). For chromogenic detection,
87

sections were stained using the avidin-biotin complex method (Vector,


Burlingame, CA) with HRP (horeseradish peroxidase)-conjugated streptavidin
and diaminobenzidine (Brown et al., 2001).
The primary antibodies were mouse anti--galactosidase (gal,
monoclonal 40-1A, 1:500, DSHB, Iowa City, IA); rabbit anti-gal (1:5000, ICN
Cappel, Aurora, OH); rat anti-gal (1:500, (Saul et al., 2008)); rat anti-BrdU
(monoclonal BU1/75, 1:100, Harlan Seralab, Indianapolis, IN); mouse
anti-calbindin (monoclonal CB-955, 1:500, Sigma, St. Louis, MO); mouse
anti-Cre (monoclonal 7.23, 1:300, Covance, Princeton, NJ); mouse anti-cyclinD1
(sc8396, 1:100, Santa Cruz Biotechology, Santa Cruz, CA); rabbit anti-GFP
(1:5000, Upstate, Lake Placid, NY); chicken anti-GFP (1:2000, Abcam,
Cambridge, MA); mouse anti-hPLAP (monoclonal 8B6, 1:250, Sigma); mouse
anti-Ki67 (monoclonal MM1, 1:25, Novocastra, Newcastle, UK); rabbit
anti-mGluR2/3 (1:200, Chemicon); goat anti-Neurod1 (sc1084, 1:50, Santa
Cruz); rabbit anti-phosphohistone H3 (1:400, Upstate, Lake Placid, NY); rabbit
anti-rhodamine (1:500, Invitrogen). Biotinylated PNA (Vector) was used at 1:250.
For simultaneous detection of BrdU (5-bromo-2-deoxyuridine) and other
markers, cryosections were fully stained with primary antibodies and lectins, and
fluorescent secondary reagents. Sections were then treated with 2.4 N HCl in
PBTx for 60-75 min at room temperature, washed, and immunostained for BrdU.
EdU (5-ethynyl-2-deoxyuridine) was detected using an azide-alkyne
cycloaddition reaction (Buck et al., 2008) and commercial reagents (Click-iT-647,
Invitrogen) after immunostaining. For EdU and BrdU co-labeling, BrdU

88

immunostaining was performed as the final step. For Ki67 immunostaining,


sections were unmasked before the blocking step by heating to 95C for 10 min
in 0.01 M citric acid.
For chromogenic detection of hPLAP activity in retina, 5-10 m
cryosections were heat-treated for 30 min in PBS with 2 mM MgCl2 at 70C and
stained with 5-bromo-4-chloro-3-indolyl phosphate (BCIP) and nitroblue
tetrazolium (NBT) substrates (Roche, Indianapolis, IN) for 1.5 hrs (Lobe et al.,
1999). To detect hPLAP activity in the brain, adult tissues from transgenic
animals were immersion-fixed in 4% PFA, 2 mM MgCl2 at 4C overnight,
heat-treated for 45 min in PBS with 2 mM MgCl2 at 70C, and embedded in 3%
agarose. Thick coronal vibratome sections (250 m) were stained for hPLAP
activity as floating slices in 24-well plates in AP buffer containing 0.01% Na
deoxycholate, 0.02% NP-40, 2 mM levamisole, and BCIP/NBT substrate
(Roche), for 5-6 hrs at room temperature. Sections were washed in PBS
containing 20 mM EDTA, dehydrated through a graded ethanol series, cleared
with BABB (1:2 benzyl alcohol : benzyl benzoate) and mounted in Permount
(Fisher Scientific, Pittsburgh, PA). Chromogenic detection of -galactosidase
(gal) activity with 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (Xgal)
substrate and in situ RNA hybridization were performed as described (Brown et
al., 2001).
Images were obtained using a Nikon Eclipse E800 epifluorescence
microscope and a SPOT digital camera. Low power images of brain sections
were captured using a Zeiss Axioimager Z1 microscope with 5X objective.

89

Confocal images were collected using a Noran OZ Laser Scanning Confocal


assembly microscope or Zeiss LSM510 Meta imaging system.
Labeling RGCs by retrograde axonal tracing
To unequivocally identify all RGCs, we performed retrograde axon labeling
with a rhodamine dextran tracer (Farah and Easter, 2005; Rachel et al., 2002).
Eyes from adult Math5>Cre; R26floxGFP mice were rapidly immersed in artificial
cerebral spinal fluid (aCSF) (von Bohlen und Halbach, 1999). The optic nerves
were transected within 1 mm of the sclerae and pressed against 4 mm cubes of
surgifoam (Ethicon, Somerville, NJ) saturated with 3% L--lysophosphatidyl
choline (LPC, Sigma) and lysine-fixable tetramethyl rhodamine dextran 3,000
MW powder (Molecular Probes, Eugene, OR). Each eye and surgifoam cube
was sealed with 1% agarose and incubated in aerated aCSF for 1 hr at room
temperature. The eyes were then incubated overnight in fresh aCSF without
surgifoam, fixed in 4% PFA for 4-6 hrs at room temperature, cryoprotected in
PBS with 10 to 30% sucrose, and frozen in OCT. For P1 mice, the same
retrograde labeling procedure was followed, except that eyes were immersed in
Hanks balanced salt solution containing calcium, magnesium and 1 mM glucose
(Gerfen et al., 2001). After labeling, eyes were fixed in 4% PFA for 1 hr. The
dissected retinas were post-fixed for 3 hrs, immunostained and flatmounted for
imaging.

90

Math5 cell cycle analysis


Retinas from Math5 +/- and Math5 -/- embryos (carrying the lacZ knock-in
allele) were co-labeled for gal, BrdU or EdU (S phase), phosphohistone H3 (M
phase, Bradbury, 1992), cyclinD1 (G1 and early S phases, Yang et al., 2006),
and Ki67 (S, G2, M and late G1 phases, Key et al., 1993). To label cells in S
phase, pregnant dams were given a single intraperitoneal injection of EdU (6.7
g/g of body mass) or BrdU (100 g/g) 30-60 min prior to harvest. To test
whether Math5+ progenitors re-enter the cell cycle, lineage-marked Math5>Cre
embryos carrying Z/AP or R26floxGFP reporters were similarly pulsed with BrdU
or EdU and their retinas co-stained for hPLAP, GFP or Cre and cell cycle
markers.
Quantitative lineage analysis
Math5+ descendants were revealed by hPLAP or GFP immunolabeling in
200 adult retinal sections. Cell types were identified by laminar position,
characteristic morphology, expression of diagnostic markers, and retrograde
axon tracing. Lineage-marked cones were distinguished from rods by co-labeling
with anti-hPLAP and PNA lectin (Blanks and Johnson, 1983). Because strong
hPLAP staining in cone pedicles obscured horizontal cell bodies, we identified
these cells using the R26floxGFP reporter and calbindin immunostaining (Peichl
and Gonzalez-Soriano, 1993). Horizontal cells were surveyed in 58 sections
from Math5>Cre; R26floxGFP mice (8 eyes). GFP-positive rhodamine-dextran
labeled RGCs and DAPI-labeled nuclei (RGCs + displaced amacrines) were
counted within the GCL in 33 fields (200X magnification) representing 8
91

Math5>Cre; R26floxGFP adult eyes. For P1 counts, the fraction of lineagelabeled RGCs was determined in retinal flatmounts from 3 eyes. The fraction of
each cell type descending from Math5+ progenitors, and the fraction of Math5+
progenitors giving rise to each cell type, were calculated based on detailed retinal
cell counts reported for adult C57BL/6J mice (Jeon et al., 1998). For lineage
tracing in the absence of Math5 gene function, labeled cells were counted in 23
fields (200X magnification) representing 6 adult eyes.
Dual reporter concordance
To assess Math5>Cre efficiency and heterogeneity among Math5+
progenitors, we crossed Math5>Cre; Z/AP mice to homozygous R26floxGFP
mice. Retinal sections from 3-4 week-old triple transgenic offspring (Math5>Cre;
Z/AP; R26floxGFP) were immunostained for GFP and hPLAP. Single- and
double-labeled cones, rods, amacrines and GCL neurons were counted in 18
fields (200X magnification) representing 4 eyes. To calculate concordance, we
divided the number of double-labeled cells by the total number of labeled cells.
Concordance was evaluated statistically using Cohens test (Cohen, 1960).
Birthdating and window labeling studies
To identify Math5 descendants exiting mitosis before P0, we performed a
cumulative BrdU labeling experiment (Miller and Nowakowski, 1988). Pregnant
dams carrying Math5>Cre; Z/AP embryos were given a single BrdU injection
(100 g/g body mass) on day E10.5 and provided with drinking water containing
500 g/mL BrdU and 1% sucrose (pH 7.0) until birth (Mayer et al., 2000). To

92

maximize labeling efficiency, water bottles were protected from light and replaced
daily. Retinal sections from 3-week-old offspring were immunostained for BrdU
and hPLAP.
To monitor how the fates of Math5+ progenitors exiting mitosis change
during development, we performed birthdating (pulse-labeling) experiments.
Pregnant dams carrying Math5>Cre; Z/AP embryos were given a single BrdU
injection (as above) on day E14.5, E15.5, E16.5 or E17.5 of development. Eyes
from 3-4 week-old mice were stained with BrdU and hPLAP antibodies, and PNA
lectin. The total number of cones (PNA+) and the number of hPLAP+ and/or
BrdU+ photoreceptors were counted in 14 central retinal fields (200X
magnification), corresponding to 3 eyes for each time-point. For birthdating
lineage-marked photoreceptors in the absence of Math5 function, we followed
the same protocol as above. We immunostained Math5>Cre; R26floxGFP;
Math5 -/- retinas for BrdU and GFP, and counted 7 fields (200X magnification)
from 2-4 eyes for each time-point. The fraction of lineage-marked and birthdated
cones was calculated directly from cell counts. The fraction of labeled rods was
estimated using a 35.2 rod-to-cone ratio for wild-type mice, based on C57BL/6
data (Jeon et al., 1998), and a 12.1 ratio for Math5 mutants (SEM = 0.8 based on
n = 5 animals, 71 fields at 200X magnification).
To determine the contribution of Math5+ cells to the early-born (EB) cohort
of neurons, we performed pulse- and window-labeling experiments at the onset
of neurogenesis. For pulse-labeling, gravid dams carrying Math5>Cre;
R26floxGFP embryos were given a single injection of EdU at day E11, and eyes

93

from the resulting pups were harvested at P1. Whole retinas were stained for
GFP and EdU, flatmounted, and imaged as confocal Z-stacks through the
ganglion cell layer. The fraction of early-born cells in the Math5 lineage (EdU+
GFP+ / EdU+) was determined from 4 eyes representing 4 mice.
For window labeling (Repka and Adler, 1992), pregnant dams carrying
Math5 +/- and -/- embryos were given EdU on day E11, as a single injection or
two injections 12 hrs apart. No difference was apparent in the extent of EdU
labeling between these schedules. Dams were then given a single injection of
BrdU on E12 and provided with BrdU in the drinking water until harvest at E12.5.
Early-born cells (EdU+ BrdU-) were counted from 3-4 embryos of each genotype,
representing 1-3 litters, and scored for gal or Brn3b immunoreactivity.
Statistical error is reported as the binomial standard deviation. Labeled fractions
were compared using Fishers exact test (Fisher, 1925).
Retinal explants and clonal analysis
Retinal explant culture and retroviral infections were performed using
established methods, which favor RGC survival (Hatakeyama and Kageyama,
2002; Wang et al., 2002b). Math5 lacZ/+ retinas were dissected from E12.5 or
E13.5 eyes, removing sclerae, pigmented epithelium (RPE) and lens tissue, and
were flattened onto 5 mm Nucleopore polycarbonate membranes (0.4 m pore
size, GE Healthcare, Piscataway, NJ). These explants were placed on Transwell
inserts (Corning) in 2-cm dishes containing neurobasal media (Invitrogen) with
1X B27 and N2 supplements, glutamine (0.4 mM), BDNF (50 ng/mL, Peprotech,
Rocky Hill, NJ), CNTF (10 ng/mL, Peprotech), penicillin (50 U/mL), streptomycin

94

(50 g/mL), and gentamicin (0.5 g/mL), and cultured at the gas-media interface
at 37C and 5% CO2.
MIG retroviral stocks (Van Parijs et al., 1999) were generated by
transfecting MSCV-IRES-GFP plasmid DNA into the Phoenix ecotropic
packaging cell line (Pear, 2001; Swift et al., 2001) and titered on NIH3T3
fibroblasts. Filtered viral preparations (~8x105 colony-forming units/mL)
containing polybrene (hexadimethrine bromide, 0.8 g/mL, Sigma Aldrich, St.
Louis, MO) were added directly to the explant surface in one drop (25 L) to
infect mitotic cells. After 2 days in vitro (DIV), half of the media was replaced
with fresh media. After 3 DIV, explants were fixed for 30 min in 4% PFA,
cryoprotected in 30% sucrose, and frozen in OCT. Thick (30 m) sections were
immunostained for gal and GFP. For each time-point, the size and composition
of clones was determined by 3-dimensional analysis of confocal Z-stacks.
Clones were defined as clusters of GFP+ cells directly apposed to each other
(within 2-3 m) and separated by at least 4 cell bodies from any other GFP+
cells. Only clones containing at least two GFP+ cells and one gal+ cell were
scored. Previous studies have shown that the average progenitor cell cycle
length is 14-16 hrs at this stage (Alexiades and Cepko, 1996; Sinitsina, 1971),
permitting 4-5 divisions during the 72 hr culture period. Accordingly, the largest
clones in each set of explants contained 8-16 cells, reflecting a minimum of 3-4
divisions in vitro.

95

Results
Math5 is transiently expressed by early retinal progenitors during or after
their terminal cell cycle
As a first step to determine the mechanism of Math5 action, we defined
the timing of Math5 expression during retinal development by quantitative PCR
(Fig. III-1A). Math5 mRNA increases rapidly at E11, peaks between E12.5 and
E14.5, and declines gradually after E14.5. This temporal profile is consistent with
RNA in situ hybridization data (Brown et al., 1998) and closely resembles
birthdating curves for RGCs (Drager, 1985; Young, 1985a). These data suggest
Math5 acts transiently during early retinal neurogenesis.
The cellular distribution of Math5 mRNA and Math5-lacZ activity across
the retinal epithelium (Brown et al., 2001) is consistent with Math5 transcription in
actively proliferating and/or postmitotic cells. Both patterns have been reported
for different bHLH genes during neurogenesis (Kageyama and Nakanishi, 1997).
Indeed, the closely related gene Math1 is expressed in mitotic cells in the
developing cerebellum (Helms et al., 2000) and in postmitotic cells in the inner
ear (Chen et al., 2002). In frog, zebrafish and chick retinas, orthologous Ath5
genes are expressed in progenitors during their last cell division (MatterSadzinski et al., 2001; Perron et al., 1998; Poggi et al., 2005).
To determine the onset of Math5 expression in individual mouse retinal
progenitors, we immunostained E13.5, E15.5, and E16.0 eyes from Math5 +/(lacZ/+) and/or Math5 -/- (lacZ/lacZ) embryos for -galactosidase (gal), the cell

96

cycle marker phosphohistone H3 (PH3, M phase), cyclin D1 (cycD1, G1/early S


phase) (Yang et al., 2006) or Ki67 (late G1, S and M phase), and the thymidine
analog EdU or BrdU (S phase) following a 30-60 minute pulse in vivo. After the
EdU pulse, a small fraction of S phase progenitor cells enter G2 and are detected
as EdU+ cycD1. In contrast, cells remaining in S phase are EdU+ cycD1+.
After careful 3-dimensional analysis of confocal Z-stack images, we observed a
small number of gal-expressing cells that had incorporated EdU at E13.5 (18 of
517 = 3.5 0.6% SD) for n = 3 sections, Fig. III-1B). These gal+ cells were
exclusively cycD1 (0 of 517, upper limit 95% CI = 0.6%), indicating that Math5 is
expressed after G1 phase at E13.5. Accordingly, gal+ PH3+ cells (M phase)
were observed at E13.5 (Fig. III-1D, (Le et al., 2006)). In contrast, in E15.5 and
E16.0 embryos (Fig. III-1C,E,G), few or no cells co-expressed gal and cell cycle
markers EdU, BrdU, cyclinD1, or PH3. The dynamics of Math5 expression thus
change during development. At early stages (<E14), some progenitors initiate
Math5 expression during the last cell cycle, whereas at later stages (>E15),
progenitors express Math5 only after terminal mitosis. Similar results were
observed in E15.5 Math5 knockout embryos (Fig. III-1F,H, (Le et al., 2006)),
demonstrating that gal+ mutant cells do not re-enter the cell cycle. In Math5 +/and -/- mice, gal+ cells span the entire retinal thickness (arrowheads in Fig. III1G,H), suggesting that radial processes associated with interkinetic nuclear
migration may persist transiently, potentially directing the migration of early postmitotic cells to their final laminar positions (Barnstable et al., 1985; McLoon and
Barnes, 1989; Snow and Robson, 1994; Watanabe et al., 1991).

97

Math5>Cre lineage marking system


We designed an expression fate-mapping system to permanently mark
lineal descendants of Math5-expressing progenitors and thereby define the range
of fates acquired by these cells. The system has two components transgenic
mice expressing Cre recombinase under strict Math5 regulatory control
(Math5>Cre) and reporter mice (Z/AP or R26floxGFP) that express a
histochemical marker (hPLAP or GFP) wherever Cre excises a loxP-flanked stop
signal.
The Math5>Cre recombinant BAC (Fig. III-2A,B) includes all likely Math5
regulatory elements (Ghiasvand et al., 2011; Hutcheson et al., 2005). We
generated nine Math5>Cre founders, each of which contains 1-5 copies of the
BAC transgene (Fig. III-S1A). Five lines were tested using Z/AP reporter mice,
which conditionally express hPLAP under control of the ubiquitous CAG promoter
(Lobe et al., 1999). Each line gave a similar staining pattern, which is consistent
with the spatiotemporal expression of Math5 mRNA (not shown). All subsequent
experiments were performed with lines 872 and 360, which contain full-length
transgene insertions, as determined by diagnostic PCR, Southern and PFGE
analysis (Fig. III-S1B).
From the onset of retinal neurogenesis (E11), Math5 mRNA is expressed
in cells near the ventricular (sclerad) neuroepithelial surface, where the majority
of progenitors undergo mitosis (Brown et al., 1998). In Math5-lacZ knock-in
mice, -galactosidase protein is expressed in a similar pattern but perdures
(Echelard et al., 1994) in the differentiating descendants of these cells, including

98

RGCs (Brown et al., 2001). In double transgenic Math5>Cre; Z/AP embryos, the
alkaline phosphatase (hPLAP) marker first appeared at E12.5 in differentiating
RGCs and the developing optic nerve (Fig. III-2D), whereas no hPLAP was
detected in control embryos carrying Z/AP alone (Fig. III-2C). At later
developmental stages, some other cell types were labeled with hPLAP (e.g.
photoreceptors at P0.5 in Fig. III-2D, arrowhead). As expected, hPLAP was only
detected in the adult retina and brain, in known Math5 RNA expression domains.
In the central nervous system, the hPLAP reporter marks neurons in the auditory
hindbrain and cerebellum (Saul et al., 2008) and reveals all known RGC
projections (Rodieck, 1998; Simpson, 1984), including those extending to the
superior colliculi, lateral geniculate bodies, suprachiasmatic nuclei, and the
accessory optic tracts (Fig. III-2E).
In the E15.5 retina, a comparison of the spatial and temporal patterns for
Math5 mRNA, Math5-lacZ and hPLAP (Fig. III-2F) is consistent with a direct role
for Math5 in RGC development and highlights the inherent time delay associated
with Cre protein synthesis, excisional activation of the Z/AP reporter, and
expression of the hPLAP enzyme (Nagy, 2000). Considering the dynamics of
retinal interkinetic nuclear migration (Baye and Link, 2008), these results suggest
there is a burst of Math5 expression in progenitors exiting the cell cycle. If Math5
is exclusively made during the last division, lineage-marked cells should never
re-enter S phase. To test this prediction, we analyzed E13.5 Math5>Cre;
R26floxGFP and E15.5 Math5>Cre; Z/AP embryos exposed to EdU or BrdU for
1 hr (Fig. III-2G,H). In E13.5 embryos after a 30 min chase, some Cre+ EdU+

99

cells were present (33 of 394 Cre+ cells = 8.4 0.4% SD for n = 3 sections) and
these were restricted to the fresh neurogenic subset (33 of 223 Cre+ GFP cells
= 14.8 1.4% SD). No GFP+ EdU+ cells were observed in the same sections (0
of 309 GFP+ cells, upper limit 95% CI = 0.9%), due to the delay in the Cre-lox
system (Fig. III-2G). Likewise, in E15.5 embryos, there was no overlap between
hPLAP activity and any cell cycle marker (Fig. III-2H). Together, these results
strongly suggest that Math5 is expressed transiently during or shortly after the
terminal cell division. Math5 lineage cells do not re-enter the cell cycle.
Quantitative Math5 lineage analysis
To reveal the fates of Math5+ progenitors, we crossed Math5>Cre mice to
Z/AP and R26floxGFP reporter strains and examined mature retinas of 3-4 week
old offspring. We observed hPLAP+ cells distributed evenly across the central
and peripheral retinas of Math5>Cre; Z/AP mice, but staining was absent in
littermates carrying the Z/AP transgene alone (Fig. III-3A,B). Because hPLAP
protein is membrane-tethered, we could identify most retinal cell types by
morphology and laminar position. As expected, RGCs were abundantly labeled.
However, we also observed significant staining among rods, cones, horizontal
and amacrine cells (Fig. III-3A,C,D). The inner plexiform layer (IPL) was
intensely labeled due to hPLAP localization in RGC and amacrine dendrites. A
thorough survey revealed rare hPLAP+ Mller glia and bipolar cells (Fig. III-3C).
Importantly, no labeling was observed in retinal cell types that have a separate
developmental origin, such as vascular endothelial cells, pericytes, microglia and

100

astrocytes, or in any other parts of the eye, including the anterior chamber and
RPE.
To systematically measure the fraction of lineage-marked retinal cells in
each class, we co-stained sections for hPLAP or GFP reporters and cell
type-specific markers. Equivalent results were obtained using Z/AP and
R26floxGFP reporters (see below) and different Math5>Cre lines (data not
shown). However, the intensity of expression varied among cell types. Z/AP is
strongly expressed in photoreceptors via the CAG promoter, whereas
R26floxGFP is weakly expressed by photoreceptors but strongly expressed by
other cell types. Consequently, we used Z/AP to count hPLAP+ and hPLAPcones (arrows in Fig. III-3D), and hPLAP+ rods (arrowheads in Fig. III-3D) in the
outer nuclear layer (ONL), and PNA lectin to distinguish cones from rods (Blanks
and Johnson, 1983). We then counted hPLAP+ bipolar cells and Mller glia (Fig.
III-3C) in the inner nuclear layer (INL) of the same sections. The labeled fraction
was calculated for each cell type using reference data for cell populations in the
adult mouse retina (Jeon et al., 1998). This fraction ranged from 31% for cones
to 1% for rods, and <0.1% for bipolar cells and Mller glia (Table III-1).
To evaluate horizontal, ganglion and amacrine cell types, we used the
R26floxGFP reporter, which co-localizes with cell type-specific markers in the
perinuclear cytoplasm. We identified horizontal cells by calbindin
immunostaining (Peichl and Gonzalez-Soriano, 1994) and their position at the
outer edge of the INL (Fig. III-3E). Twenty-nine percent of horizontals were
GFP+ (Table III-1). This value is significantly lower than that reported by Yang et

101

al. (2003), but consistent with horizontal cell labeling data of Feng et al. (2010, cf.
Suppl. Fig. 3E) obtained using a Math5-Cre knock-in allele. RGCs were
distinguished from displaced amacrine cells (Hayden et al., 1980; Perry and
Walker, 1980) by retrograde rhodamine-dextran tracing of optic nerve axons.
Forty-three percent of neurons in the ganglion cell layer (GCL) were labeled with
rhodamine in this experiment (arrows, Fig. III-3F), in close accord with previous
data (Jeon et al., 1998). All other cells in the GCL were scored as displaced
amacrines (arrowheads, Fig. III-3F). The frequency of GFP labeling in the adult
retina was 55% for ganglion cells, 28% for displaced amacrines, and 9% for INL
amacrines (Table III-1). To evaluate the Math5 lineage fraction prior to the
normal period of RGC culling (Galli-Resta and Ensini, 1996), we performed a
similar analysis in early postnatal retinas, limited to the GCL (Fig. III-3G). The
fraction of GFP+ ganglion cells in P1 retinas (53 1%, n = 3, 948/1777 cells) was
similar to that observed in the adult (55 2%, Table III-1, Fishers exact test P =
0.3).
A clear pattern emerges from these data. First, Math5+ progenitors have
the potential to develop into all seven major retinal cell types. Second, the
distribution of Math5+ descendents differs from the retina as a whole (Fig. III-3I,
2 test with df = 7, P < 0.0001). Third, the labeling fraction of each cell type
(Table III-1) decreases according to the birth order (Carter-Dawson and LaVail,
1979; Rapaport et al., 2004; Sidman, 1961; Young, 1985a). Early-born cell types
RGCs, cones, horizontal cells and displaced amacrines frequently descend
from Math5+ progenitors, whereas late-born bipolar and Mller glial cells rarely

102

derive from Math5+ progenitors. INL amacrines are born during the middle
phase of retinal development, prior to the peak of rod births, and these cell types
have an intermediate labeling fraction. We estimate that 3% of adult retinal cells
descend from Math5+ progenitors (Table III-1). Fourth, only one in nine Math5+
descendants is a ganglion cell (11%). Because RGCs represent ~0.5% of
neuroretinal cells in adult mice (Jeon et al., 1998) and Math5 status does not
affect RGC survival between P1 and adulthood, Math5 descendants are 50-fold
more likely on average to develop as RGCs than the remaining neuroretinal
population (approx. 1 in 500). Fifth, 45% of ganglion cells are not marked by the
Math5>Cre transgene, suggesting the possibility of a substantial
Math5-independent RGC subpopulation. Although the fraction of GCL neurons
labeled by Math5>Cre (40%, Table III-I) approximates the RGC fraction (43%),
this value includes both RGC (24%) and displaced amacrine cell types (16%).
The fate of Math5 mutant cells
In mutant mice, the Math5 transcription unit is active, expressing lacZ
mRNA, but lineage-marked progenitors are blocked from developing as RGCs.
To determine the fates of these cells, we examined retinas from adult Math5 -/mice carrying Z/AP and Math5>Cre transgenes (Fig. III-3H). The extent of
hPLAP labeling in the mutant retina was roughly similar to wild-type (Fig. III-3A).
However, the fate profile within the Math5 lineage was different (2 test with df =
7, P < 0.0001). First, RGCs were absent, as expected, decreasing the amount of
IPL staining. Second, there was an obvious increase in late-born cell types
among the hPLAP+ neurons (Table III-S1). For example, rod photoreceptors

103

increase from 32% to 40% of the Math5 lineage. Labeled bipolar cells and Mller
glia were visible in most low power fields (200X magnification) of mutant mice,
but were difficult to find in wild-type Math5>Cre; Z/AP retinas (Table III-1),
consistent with results observed by Feng et al. (2010). This effect is more
striking if one considers that the total number of rods, bipolar cells and Mller glia
are decreased in Math5 mutants (Brown et al., 2001; Brzezinski et al., 2005). In
Math5 mutant mice, the cohort of progenitors expressing Math5>Cre does not
differentiate into RGCs, but retains competence to develop into any of the
remaining principal cell types.
Math5+ progenitors have equivalent Cre activity
Only a small fraction (11%) of the Math5 lineage develops into RGCs. In
principle, the Math5+ population may be heterogeneous, such that one group of
progenitors expresses high levels of Math5>Cre and develops as RGCs, while a
second group expresses low levels of Math5>Cre and adopts other fates (Fig. III4A). In this model, the low-level multi-lineage Math5>Cre expression could
represent priming (Hu et al., 1997) of the Math5 gene, or leaky transgene
expression, an over-reporting artifact that is not biologically meaningful
(Dymecki et al., 2002). Alternatively, all Math5+ progenitors may express
equivalent levels of Math5>Cre (Fig. III-4B), consistent with a permissive role for
Math5 in retinal development.
To test these alternatives, we examined retinas from triple transgenic
(Math5>Cre; Z/AP; R26floxGFP) adult mice, using the concordance of hPLAP
and GFP labeling in Math5 descendants as an indirect measure of Cre activity

104

(Fig. III-4C). In this experiment, we assume that the probability of reporter


activation in a particular cell depends on the concentration and stability of
intracellular Cre protein, and that neither reporter is saturated at the Cre levels
under investigation. Progenitors with strong Cre expression are expected to
activate both reporters, while progenitors with weak Cre expression may
randomly activate one reporter, Z/AP or R26floxGFP, at a low frequency (Fig. III4A,B). If these events occur independently with equal probability (), then the
joint probability of observing both reporters in a single cell (expected
concordance) should equal 2/(2-2), where 2 is the fraction of cells that
activate both reporters and 2-2 is the fraction of cells that activate at least one
reporter. The observed concordance was uniformly high (~80%) for rods, cones,
INL amacrines and GCL neurons, and much greater than expected by chance
(Cohens > 0.7, Fig. III-4D, Table III-S2). Thus, the labeling of non-RGC cell
types cannot be attributed to differential or leaky Math5>Cre expression.
The fate of the Math5+ progenitor population changes over time
The discovery that some rods, bipolars and Mller glia descend from
Math5>Cre progenitors (Table III-1) is somewhat surprising because the vast
majority of these cell types undergo terminal mitosis (Rapaport et al., 2004;
Young, 1985a) after the temporal envelope of Math5 mRNA expression (Fig. III1A). To explain these findings, we performed a cumulative labeling experiment,
in which Math5>Cre; Z/AP embryos were continuously exposed to BrdU from
E10.5 until P0 and analyzed at P21. Retinal progenitors that exit mitosis before
P0 should be heavily BrdU-labeled, whereas those that continue to divide after
105

P0 should be weakly labeled. We found that essentially all hPLAP+ cells in the
central retina were heavily labeled with BrdU (98.8%), including rods
(arrowheads in Fig. III-5A), cones (arrows in Fig. III-5A), and INL and GCL
neurons (Table III-S3). Therefore, Math5+ rods, bipolars and Mller glia are born
at the leading edge of birthdating curves for these late cell types.
The fate profile of neurogenic cells emerging from the RPC population is
known to change over time, in response to intrinsic factors and environmental
signals (Livesey and Cepko, 2001; Rapaport et al., 2004; Young, 1985a). This
can occur through alterations in the fate bias of individual cells or the composition
of the RPC pool (heterogeneity). In principle, the Math5+ cohort may behave
similarly. The fate profile of these cells may be intrinsically programmed, or it
may vary depending on the time that an individual RPC exits mitosis and initiates
Math5 transcription. To test these alternatives, we compared the fates of Math5+
progenitors born on different days. Math5>Cre; Z/AP embryos were exposed to a
pulse of BrdU on E14.5, E15.5, E16.5 or E17.5 and their adult retinas were
examined by hPLAP, PNA, and BrdU staining (Fig. III-5B). A variety of
lineage-marked cell types were born on each of these days, including RGCs,
rods, cones, amacrines and horizontal cells, as well as rare late cell types
(arrowhead in Fig. III-5B). For quantitative analysis, we focused on
photoreceptors, which are relatively numerous and could be directly compared
within the ONL. At each time-point, we determined the fraction of hPLAP+ and
heavily BrdU+ rods and cones in the central retina (arrows in Fig. III-5B, Table IIIS4). The fraction of photoreceptors (rods plus cones) that were derived from

106

Math5+ progenitors decreased between E14.5 and E17.5, from 20.6% to 4.3%
(Fig. III-S2, Table III-S4), in parallel with the decrease in the total number of
Math5+ cells.
The fate of the Math5+ cell population also changed significantly between
E14.5 and E17.5, together with the retina as a whole. Math5+ cells born on
E14.5 were >2 times as likely to develop into cones as compared to rods (136 vs.
64), whereas those born on E17.5 were >60 times as likely to develop into rods
as compared to cones (122 vs. 2, Table III-S4). The fates of progenitors inside
and outside the Math5 lineage shifted in parallel, as shown by plots of the coneto-rod ratio (Fig. III-5C), derived from birthdating curves (Fig. III-S2). This shift is
primarily determined by the overall decrease in cone births during this interval.
Math5+ cells appear to follow the same fate trajectory as other progenitors.
However, the ratio curves are displaced by one-half day. In comparison to other
neurogenic cells (hPLAP-) exiting mitosis on the same day in the same retinal
environment, Math5+ progenitors (hPLAP+) were three times more likely to
develop into cones. Surprisingly, similar results were obtained in the absence of
Math5 function, in mutant embryos carrying R26floxGFP and Math5>Cre
transgenes (Fig. III-5D).
These findings support three conclusions. First, the fate profile of Math5+
cells changes over time, similar to that of other retinal progenitors. Second, the
fate bias of Math5+ cells extends beyond RGC specification, influencing the
choice among alternative fates (e.g. cone vs. rod). Third, the bias among

107

alternative fates is independent of Math5 action, suggesting that upstream or


parallel factors are responsible.
Math5 expression in early-born retinal cells
Because Math5 expression is closely correlated with the onset of retinal
neurogenesis (~E11.5) (Hufnagel et al., 2010) and is essential for specification of
the earliest born cell type, RGCs (Brown et al., 2001; Wang et al., 2001), we
expected that most or all early-born retinal cells would express Math5 and adopt
RGC fates. To test this hypothesis, we performed a series of window-labeling
experiments. Embryos were sequentially exposed to EdU at E11 and BrdU at
E12 (Fig. III-6A). In this paradigm, cells incorporate EdU if they are in S phase at
E11. Because the average cycle length at this stage is less than 24 hrs
(Sinitsina, 1971), EdU+ BrdU+ cells scored at E12.5 are interpreted as RPCs that
underwent one additional division (and S phase). In contrast, EdU+ BrdU- cells
define the early-born (EB) cohort. These cells were in S phase at E11, but exited
the cell cycle before E12.5.
To evaluate RGCs within the EB cohort, we counted the fraction of EdU+
BrdU cells that were Brn3b+ (RGCs) in Math5 heterozygous and mutant mice
(Fig. III-6B). In Math5 +/- embryos, 75% of the EB cohort expressed Brn3b,
confirming that RGCs are the predominant first-born cell type (Farah and Easter,
2005; Rachel et al., 2002). The abundance of EdU+ BrdU- cells was similar in
Math5 -/- and +/- embryos (5.7 vs. 7.1 per 0.001 mm2 field, respectively) and
comparable to previous birthdating results (Le et al., 2006). However, in Math5
mutant embryos, only 6% EdU+ BrdU- cells expressed Brn3b. This was

108

expected from the deficiency of RGCs in these mice, and confirms that the loss
of RGCs is an early event. We next determined the fraction of EdU+ BrdU- cells
that expressed Math5, using the lacZ allele (gal) as a short-term lineage tracer
(Wang et al., 1999). Surprisingly, only 20% of EdU+ BrdU cells were gal+, in
both Math5 +/- and Math5 -/- mice (Fig. III-6C). To independently test this result,
we exposed Math5>Cre; R26floxGFP embryos to a single pulse of EdU at E11,
harvested their retinas at P1, and determined that 28% of strongly EdU+ cells in
the GCL were GFP+ (Fig. III-6D,E). As a third test, we evaluated retinas from
early Math5-lacZ/+ embryos for coexpression of lacZ and Brn3b. The fraction of
gal+ RGCs was relatively low at E11.5, consistent with the EB analysis, but
increased from 20% to 60% between E11.5 and E13.5 (Fig. III-7). Taken
together, the results from these three experiments suggest that Math5 is
expressed by a subset of early neurogenic cells, and that only a fraction of
Brn3b+ RGCs generated at E11-13 derive from the Math5+ cohort.
The Math5-independent early-born cells may express other proneural
bHLH transcription factors in the Atonal family, such as Neurod1 or Neurog2. At
E11.5, Neurod1 was detected in a pattern that partially overlaps Math5-lacZ (Fig.
III-S3), consistent with mRNA in situ hybridization data (Hufnagel et al., 2010). A
similar overlap has been noted later in development (Kiyama et al., 2011; Le et
al., 2006). This may explain the small number of early-born Brn3b+ RGCs
present in Math5 -/- mice (Fig. III-6B), as Neurod1 can partially substitute for
Math5 function in RGC fate specification (Mao et al., 2008b).

109

Symmetry of Math5 expression in marked retroviral clones


During nervous system development, the mode of progenitor cell divisions
changes over time (Gotz and Huttner, 2005; Huttner and Kosodo, 2005; Lu et al.,
2000). Prior to neurogenesis, cell divisions predominantly follow the symmetric
self-renewing mode (P-P), which expands the progenitor pool. During early
neurogenesis, an asymmetric mode is frequently used to generate one mitotic
daughter and one differentiating neuron (P-N). During late neurogenesis, most
progenitors undergo a symmetric neurogenic mode of division (N-N), in which
both daughters permanently exit the cell cycle. The fates adopted by neuronal
daughters may also be symmetric (Na-Na) or asymmetric (Na-Nb). In zebrafish,
retinal progenitors expressing ath5-GFP undergo terminal neurogenic cell
divisions (Poggi et al., 2005). These are symmetric with respect to ath5-GFP
expression (NGFP-NGFP), but the daughters may have different cell fates,
depending on the retinal environment.
To examine the mode of RPC division giving rise to Math5+ daughter cells
in mice, we infected retinal explants from E12.5 or E13.5 Math5-lacZ/+ embryos
with MSCV-IRES-GFP (MIG) retrovirus at low density to mark independent GFP+
clones. After culturing explants for 3 days in vitro (DIV), we immunostained 30
m cryosections for GFP and gal (Math5-lacZ), and determined the size and
composition of clones containing at least one gal+ cell (Ngal) (Fig. III-8A).
These GFP+ clones ranged from 1 to 16 cells. We then focused our analysis on
small clones (2-4 cell) containing 1 gal+ cell, as these were most informative
for symmetry of gal+ expression. These clones are likely to represent terminal

110

lineages, given their small size and time in culture. Indeed, all cells in these
small clones were postmitotic, as judged by expression of the cell cycle inhibitor
p27Kip1 (Dyer and Cepko, 2001a) (data not shown). Among 23 clones scored,
we observed both symmetric (Ngal-Ngal) (Fig. III-8B,C) and asymmetric (N-Ngal,
or possibly P-Ngal) (Fig. III-8D) patterns of Math5 expression. Of 23 informative
neurogenic divisions, 13 (57%) were symmetric with respect to Math5 expression
and 10 were asymmetric (Fig. III-8E). The fraction of symmetric divisions did not
differ significantly between the E12.5 and E13.5 explant time-points (0.64 vs.
0.50 respectively, Fishers exact test P = 0.7). Although few symmetric terminal
divisions are expected at this age in the retina as a whole, the high frequency
observed among the Math5+ cohort (Ngal-Ngal) confirms that early progenitors
are capable of N-N divisions in mice as in zebrafish (Poggi et al., 2005). Unlike
zebrafish, neurogenic divisions can be asymmetric with respect to Math5
expression in mice. These findings confirm that many retinal progenitors express
Math5 after terminal M phase.

Discussion

Math5>Cre transgene recapitulates endogenous Math5 expression


We believe that the Math5>Cre transgene is expressed in the same
pattern as endogenous Math5 mRNA for several reasons. First, the BAC
transgenes that we examined are intact and contain >100 kb flanking Math5
genomic DNA on both sides of the Cre cassette, while core regulatory elements
for Math5 retinal expression are located within 25 kb of the transcriptional start

111

site (Ghiasvand et al., 2011; Hutcheson et al., 2005). Second, the


spatiotemporal pattern of Z/AP activation is congruent with Math5 mRNA and
Math5-lacZ expression during retinal development. Third, all lineage-marked
adult cells are born during the normal period of Math5 expression, including rare
late cell types. Fourth, similar results were observed with independent BAC
transgenic lines, suggesting that chromosomal position effects are minimal or
nonexistent. Apart from the retina and RGC projections, hPLAP staining was
only noted in the cerebellum and in bushy cells of the cochlear nucleus, tissues
that are known to express Math5 mRNA (Saul et al., 2008). Fifth, the dual
reporter concordance experiment provides no evidence for leaky or ectopic Cre
expression. Sixth, a similar overall retinal pattern has been observed using a
targeted Cre insertion (knock-in allele) in the Math5 locus (Feng et al., 2010;
Yang et al., 2003).
Our quantitative analysis significantly extends these previous studies and
allows us to reach different conclusions regarding: [1] the size of the
Math5-independent cohort of RGCs, [2] the relationship between Math5
expression and cell cycle exit, [3] the role of Math5 in determining non-RGC
fates, and [4] the diversity of cell types within the Math5 lineage.
Math5>Cre does not mark all RGCs
Since Math5 is necessary for RGC development, and functions as an
intracellular factor, we expected all ganglion cells to be labeled by Cre, as
descendants of Math5+ progenitors. However, after carefully excluding displaced
amacrine cells, we found that only 55% of RGCs were marked by the Math5>Cre

112

transgene. A similar fraction of RGCs is likely to be labeled by the Math5-Cre


knock-in allele (Feng et al., 2010, cf. Suppl. Fig. 5D), although this finding was
not originally appreciated (Yang et al., 2003). These Math5 descendants project
to all known target sites for RGCs in the brain (Fig. III-2E), suggesting that they
represent the ganglion cell population as a whole. There are two possible
explanations for the incomplete marking of RGCs: [1] inefficiency of the Cre-lox
system, and [2] the existence of a sizeable Math5-independent population of
RGCs.
In principle, inefficient Cre reporting may account for a substantial fraction
of unlabeled RGCs in the birthdating and lineage tracing experiments. RGCs
descending from Math5+ precursors may escape detection for two reasons.
First, the absolute level or duration of Cre expression in individual cells may not
be sufficient to catalyze robust recombination. The Cre polypeptide must
assemble into tetramers for enzymatic activity and has a short half-life in
mammalian cells (Nagy, 2000). Generally speaking, Cre transgenes that are
continuously active in differentiated cells are expected to be more efficient than
those that are made briefly in a progenitor population. Math5 is transiently
expressed during retinal development (Fig. III-1A) and may be transcribed for
only a few hours in individual cells (Fig. III-2F) (Fu et al., 2009). Consequently, in
the dual concordance experiment, Math5>Cre activated only one reporter in 20%
of marked cells (Fig. III-4). However, because concordance was relatively high
among all cell types (Table III-S3), this effect cannot fully explain the incomplete
labeling of RGCs. Second, some cells may epigenetically silence the Math5>Cre

113

transgene or the Math5-lacZ allele, or may be otherwise globally resistant to Cre


recombination. Indeed, we have observed rare mice with reduced or elevated
RGC labeling (data not shown). In these retinas, the extent of labeling varied
coordinately across different cell types, consistent with a clonal epigenetic effect.
A similar variation has been noted among Tie1>Cre transgenic mice in the
efficiency of endothelial cell labeling (Enge et al., 2002). Nonetheless, among
the vast majority of Math5>Cre retinas, there was relatively little variation in the
RGC labeling fraction (Table III-1). Taken together, Cre inefficiency and
epigenetic silencing are unlikely to explain the incomplete labeling of RGCs that
we observed.
Alternatively, a subset of RGCs may develop independently of Math5.
Detailed analysis of Math5 -/- retinas has revealed a small population of widely
dispersed ganglion cells, approximately 4% of wild-type, that survive to adulthood
(Lin et al., 2004) and may project to the superior colliculi and laterial geniculate
nuclei (Triplett et al., 2011). Moreover, recent data show that a related bHLH
factor, Neurod1, can partially substitute for Math5 and allow RGC development
(Mao et al., 2008b). Indeed, we observed that fewer early-born cells express
Math5-lacZ than Brn3b (Fig. III-6B,C) and that many Brn3b+ RGCs at E11-E13
do not express Math5-lacZ (Fig. III-7). A subset of nascent ganglion cells may
develop from Neurod1+ precursors (Fig. III-S3), without Math5. Consistent with
this idea, mutant mice lacking both factors mice have even fewer RGCs than
Math5 -/- mice (Kiyama et al., 2011).

114

The fraction of unmarked RGCs (~45%, Table III-1) is 10-fold greater than
the number of RGCs that survive in Math5 -/- mice (~4%)(Lin et al., 2004). Apart
from Cre inefficiency (noted above), there are two possible explanations for this
discrepancy. First, RGCs derived from Math5+ progenitors may have a survival
advantage during neonatal period (P0-P10) of ganglion cell apoptosis (Young,
1984). However, the deficiency of Math5-independent RGCs in Math5 mutants
was clearly evident early in retinal histogenesis, at E12.5 (Fig. III-6B,C), well
before the neonatal period of RGC culling. In addition, the fraction of Math5+
RGCs in P1 and adult retinas was the same, making this mechanism unlikely.
Second, Math5 lineage cells may have a substantial non-autonomous role in
RGC fate specification or early differentiation. These cells may represent
pioneering neurons (Pittman et al., 2008; Raper and Mason, 2010), which
promote axon pathfinding and fasciculation within the retina (Erskine and
Herrera, 2007; Oster et al., 2004) and survival of Math5-independent RGCs. In
the absence of Math5, cells in the inner retina undergo apoptosis during
midgestation and surviving RGCs have severe pathfinding defects (Feng et al.,
2010; Kiyama et al., 2011; Moshiri et al., 2008); Chapter V). Most likely, Math5+
progenitors may favor the formation or survival of other RGCs by para- or
juxtacrine signaling. Further work is needed to clarify molecular differences
between the Math5+ cohort and other cells in the early retina.
Math5 is made by progenitors exiting the cell cycle
We have determined the precise relationship between onset of Math5
expression and the cell cycle status of retinal progenitors (Fig. III-9A). At early

115

stages (<E14), Math5-lacZ was detected in some G2/M phase progenitors but
was otherwise present only in non-proliferating cells. Based on the length of G2
phase (~ 2 hrs) (Sinitsina, 1971) and our analysis of retinal cell cycle kinetics in
E13.5 Math5>Cre; R26floxGFP embryos, following a 30 min EdU pulse (Fig. III2G), we conclude that at least 15% (and up to 60%) of newly Math5+ cells (Cre+
GFP-) initiate expression before terminal M phase. During later stages (>E15),
Math5 was exclusively expressed in post-mitotic cells. Math5 lineage cells did
not re-enter the cell cycle at any stage, regardless of the Math5 genotype. This
comprehensive analysis reconciles previous disparate observations regarding
the timing of Math5 expression (Brown et al., 1998; Le et al., 2006; Yang et al.,
2003), including RNA profiling of single retinal cells (Trimarchi et al., 2008). In
recent studies, an HA epitope-tagged Math5 allele was expressed with similar
kinetics in early E12.5-E14.5 embryos, but was detected in more S, G2, and M
phase cells than our Math5-lacZ allele (Feng et al., 2010; Kiyama et al., 2011;
Wu et al., 2012). This is comparable to zebrafish, where ath5-GFP expression
initiates during terminal S/G2 (Poggi et al., 2005), and is consistent with results
obtained in frog and chick (Kay et al., 2001; Matter-Sadzinski et al., 2001; Perron
et al., 1998; Poggi et al., 2005).
The variable timing of Math5 expression was supported by detailed clonal
analysis. We observed symmetric Math5-lacZ expression in 13 of 23 informative
divisions (56%, NMath5-NMath5) and asymmetric expression in the remaining clones
(P/N-NMath5). This frequency is convergent with cell cycle kinetic data discussed
above. Together, these findings suggest that early progenitors giving rise to

116

Math5+ cells are heterogeneous in their intrinsic properties and/or responses to


the retinal microenvironment. By comparison, zebrafish ath5 is expressed
symmetrically in terminal neurogenic divisions (Path5 Nath5-Nath5), but resulting
daughters often adopt different fates (Poggi et al., 2005). This difference may be
correlated with the accelerated pace of retinal neurogenesis in zebrafish
compared to mice. Further studies are needed to determine how the timing of
Math5 expression influences the fate choice of daughter cells in mice.
Math5 is unlikely to autonomously regulate the decision to exit the cell
cycle for two reasons. First, it is variably expressed during or after the terminal
division (Figs. 1B-D, 2G,H). Second, Math5 lineage cells exhibit similar lacZ and
Cre reporter expression kinetics in mutant and wild-type mice (Fig. III-1E-H).
Instead, this binary choice must be made upstream or in parallel with Math5
transcription. However, Math5 may affect progenitor cycling indirectly. For
example, differentiating RGCs secrete sonic hedgehog (Shh), which acts as a
mitogen for RPCs and promotes rod and Mller glial fates (Jensen and Wallace,
1997; Levine et al., 1997; Wang et al., 2005; Yu et al., 2006). Accordingly,
Math5 mutants have thinner retinas with fewer rods and glia compared to
wild-type mice (Brown et al., 2001), which may reflect a general loss of late-born
cells. In zebrafish, ath5 and syu (Shh) mutants exhibit similar defects in retinal
cell mitosis, involving a delay in switching polarity of division, from centralperipheral (proliferative) to circumferential (neurogenic) modes (Das et al., 2003).
This likewise suggests that nascent RGCs, not the ath5 product per se, affect
progenitor cell cycle dynamics.

117

Math5 establishes an RGC competence state


The expression fate mapping (Fig. III-3) and dual concordance (Fig. III-4)
experiments support six conclusions. First, only a small fraction (3%) of the
retina derives from Math5+ progenitors. Second, Math5+ progenitors are
multipotent. They retain the potential to generate all seven major retinal cell
types. Third, Math5+ progenitors contribute differentially to each cell type. The
labeling frequency for a given cell type depends on the histogenic birth order and
the temporal expression profile for Math5 (Fig. III-9B). Similar overall results
were observed in a previous Math5 lineage study (Feng et al., 2010; Yang et al.,
2003). However, bipolars and Mller glia were not identified in the wild-type
Math5 lineage, presumably because fewer cells were sampled. Fourth, Math5+
progenitors express uniform levels of Math5 and may represent a developmental
equivalence group, similar to ato+ progenitors in the fly eye imaginal disc
(Dokucu et al., 1996). Fifth, the fates of the Math5+ and Math5- populations
change over time with parallel trajectories. The Math5 lineage cells are biased in
their selection of non-RGC fates, compared to other neurogenic cells in the same
environment, but this difference does not depend on Math5 activity. Sixth, the
diversity of retinal cell fates within the Math5 lineage was similar in mutant and
wild-type mice, apart from the deficiency of RGCs (Fig. III-9C). However, there
were modest increases in the labeling fraction of rods, bipolar cells and Mller
glia (Fig. III-3I). The most parsimonious explanation for these quantitative effects
is a difference between wild-type and mutant mice in the fractional distribution of
cell types, creating a denominator problem. As noted above, Math5 -/- retinas

118

have significantly fewer late-born cells, presumably due to loss of Shh. Because
all Math5 lineage cells retain early birthdates in these retinas (Fig. III-6C, Fig. IIIS2), this cohort appears expanded and skewed toward late cell fates. Taken
together, we conclude Math5 does not directly control the acquisition of multiple
retinal cell fates (Feng et al., 2010). Instead, Math5 has an active role in RGC
fate specification, as a competence (permissive) factor, and a passive or minor
role in the selection of alternative (non-RGC) fates.
Mechanisms of fate determination in the mouse retina
Retinal cell fate choice, differentiation and survival are jointly controlled by
intrinsic and extrinsic factors (Livesey and Cepko, 2001). As a nuclear bHLH
protein, Math5 is an intrinsic factor. It is necessary but not sufficient for RGC
development. During retinogenesis, nine-fold more Math5+ cells are produced
than develop into RGCs (Fig. III-3, Table III-1). These cells have a different fate
bias than other neurogenic cells in the same environment (Fig. III-5). This
property is conferred upstream of Math5. The development of RGCs from
Math5+ cells may require the presence of positive cofactors or the absence of
inhibitors. Soluble factors and cell-cell signaling are known to negatively regulate
RGC genesis, including factors secreted by nascent RGCs (Austin et al., 1995;
Belliveau and Cepko, 1999; Waid and McLoon, 1998; Zhang and Yang, 2001),
and these may act on Math5+ cells. Together, our data suggest that the Math5+
cohort is influenced by intrinsic and extrinsic factors.
Our finding that Math5+ progenitors born on the same day can give rise to
early or late cell types (Fig. III-5) is consistent with a progressive restriction

119

model for retinal neurogenesis, in which the progenitor pool is initially multipotent,
but gradually loses competence to form early cell types (Pearson and Doe, 2003;
Shen et al., 2006). This model is favored by heterochronic co-culture
experiments (Reh, 1992; Watanabe and Raff, 1990) and Ascl1 (Mash1) lineage
analysis. Mouse Ascl1+ progenitors form all retinal cell types except RGCs
(Brzezinski et al., 2011) and may represent the first competence-restricted state.
However, our results are also consistent with a temporal restriction model, in
which progenitors proceed unidirectionally in time through a relatively fixed series
of competence states (Wong and Rapaport, 2009).
The reservoir of neurogenic cells that are competent to form RGCs greatly
exceeds the final number. Likewise, the period of RGC competence extends
beyond the normal time envelope for RGC births in rat and chick (James et al.,
2003; Silva et al., 2003). This excess capacity, which includes Math5+ and
Math5- cells, and the fate plasticity of Math5+ cells may serve to enhance the
robustness of RGC development and ensure an appropriate histotypic profile in
the mammalian retina.
Acknowledgements
The authors are grateful to Thom Saunders, Maggie van Keuren and the
UM transgenic animal model core for generating the BAC transgenic animals; to
Sue Tarl and Dellaney Rudolph for technical support; to Nadean Brown for in
situ hybridization data; to Nathaniel Heintz for BAC targeting plasmids and
strains; to Sally Camper for the nlsCre plasmid; and to Sean Morrison for the
MIG retroviral construct. The authors thank Chris Edwards, the UM microscopy

120

and image analysis laboratory staff, Rafal Farjo and Mohammad Farah for
technical advice. The authors are grateful to Nadean Brown, Chris Chou, David
Turner, Matt Wilken, Julia Pollak, Anna La Torre, and Yumi Ueki for valuable
discussions and critical reading of the manuscript. This research was funded by
National Institutes of Health (NIH) R01 grant EY14259 (TG). JAB and LP were
supported by NIH T32 grants EY13934 (JAB, LP), GM07544 (JAB), and
GM07863 (LP).

121

Figure III-1. Math5 is expressed by early retinal progenitors during or shortly


following their terminal cell cycle. (A) Time course of Math5 mRNA expression in
developing eyes. Math5 mRNA levels peak at E14, with a profile that resembles
RGC birthdating curves (Young 1985; Rapaport et al. 2004). (B-C) Sections from
E13.5 (B) or E16.0 (C) Math5 +/- embryos co-stained for gal (Math5-lacZ allele),
EdU (following a 30 min chase), and cyclin D1 (marks G1/early S phase). Upper
and lower panels show single- and double-labeled confocal projection images of
10 (B) or 3 (C) 1-m optical slices. Insets show a gal+ cell in G2 (EdU+
cyclinD1). At E13.5, some gal+ EdU+ cells are present (arrows, 18 of 517
gal+ cells), but none are cyclinD1+ (0 of 517). At E16.0, few or no gal+ cells
are EdU+ or cycD1+. (D-H) Retinal sections from Math5 +/- (D, E, G) and Math5
-/- (E, G) mice co-stained for gal and cell cycle markers. Math5-lacZ is
occasionally co-expressed with M-phase marker PH3 at E13.5 (arrow in D,
inset), but does not overlap with PH3 (arrows in E,F) or BrdU (1 hr chase)
(arrows in G,H) at E15.5. Therefore, Math5-lacZ expression initiates during
terminal G2 phase at E13.5, but after terminal M phase at E15.5 and E16.0 in
both Math5 +/- and -/- retinas. M5, Math5; gal, E. coli -galactosidase; cycD1,
cyclinD1; PH3, phosphohistone H3. Scale bar, 50 m.

122

123

Figure III-2. Construction and expression of the Math5>Cre transgene. (A)


Math5>Cre BACs were generated in E. coli by a two-step homologous
recombination procedure. The single-exon Math5 open reading frame was
precisely replaced with a nlsCre-pA cassette, using A and B homology arms
derived from 5 (red box) and 3 (cyan box) UTRs. The pLD53 shuttle vector
contains recA recombinase, positive (amp) and negative (sacB) selection
cassettes, and the R6K origin of replication. The pBACe3.6 vector (gray box)
contains the chloramphenicol resistance gene (Cm) and P1 origin (ori). (B)
Confirmation of recombinant BAC structure using diagnostic (dx) PCRs 1-6
indicated in panel A (assembled from multiple gels). (C-F) Developmental
expression pattern. Math5>Cre mice were crossed to mice carrying the Z/AP
transgene, which permanently reports Cre activity. Alkaline phosphatase
(hPLAP)-positive RGCs (purple, arrows) are first observed at E12.5 in double
transgenic embryos (D), while control littermates (C) containing only the Z/AP
transgene are negative. hPLAP activity increases from E13.5 to E15.5 as RGCs
develop and form the optic nerve. By P0.5, hPLAP activity is abundant in RGCs
(arrow) and can be detected in some photoreceptors (arrowhead); however most
of the retina is unlabeled. (E) Composite images of 250 m coronal vibratome
sections through the adult thalamus and optic chiasm (inset) show the axonal
projections of RGCs derived from Math5+ precursors. Lower panels show
sections through the accesory optic system (left) and superior colliculus (right).
Labeled RGCs project to all major ganglion cell target sites in the CNS. No
significant staining was observed in the cerebral cortex or hippocampus. (F)
Kinetics of Math5 expression in the E15.5 retina. Math5 mRNA (in situ
hybridization) is expressed in retinal progenitor cells, while the cytoplasmic
Math5-lacZ knock-in allele labels progenitors (sclerad) and developing RGCs
(vitread), which have recently transcribed Math5 (-galactosidase activity).
hPLAP activity in Math5>Cre; Z/AP mice is localized to developing RGCs on the
vitread side of the retinal epithelium. These patterns demonstrate the
spatiotemporal progression of Math5 expression, if one considers the perdurance
of -galactosidase, the delay associated with Cre excision, and interkinetic
nuclear migration. Math5 is expressed transiently in progenitors that become
RGCs. (G-H) Math5>Cre retinas co-stained for lineage tracers and cell cycle
markers. Cre is expressed with the same kinetics as Math5, whereas GFP or
hPLAP reporters are expressed with a delay. (G) At E13.5, Cre+ EdU+ cells are
present (30 min chase, arrows, 33 of 394 Cre+ cells), but no GFP+ EdU+ cells
are observed (0 of 309 GFP+ cells). (H) At E15.5, no hPLAP+ cells (arrows) are
co-labeled with BrdU (1 hr chase), PH3, or Ki67 (marks late G1 through M
phase). Together, these results indicate that cells in the Math5 lineage do not reenter the cell cycle. pA, polyadenylation signal; SC, superior colliculus; BSC,
brachium of the superior colliculus; LGBd, lateral geniculate body, pars dorsalis;
LGBv, lateral geniculate body, pars ventralis; LTN, lateral terminal nucleus ; AOT,
accessory optic tract; OT, optic tract; TV, third ventricle; SCN, suprachiasmatic
nucleus; OC, optic chiasm; H, hippocampus; PN, pons. Scale bars, 100 m in
C-F; 50 m in G-H.
124

125

Figure III-3. Math5+ progenitors contribute differentially to all retinal cell types.
Math5>Cre mice were crossed to Z/AP (A-D) or R26floxGFP reporter (E-G)
strains. (A) In Math5>Cre; Z/AP mice, hPLAP+ descendants of Math5+
progenitors represent 3% of adult retinal cells (see Table III-1) and are present in
every cell layer. (B) Z/AP-only control retinas have no hPLAP activity. (C)
Math5+ descendants, detected by hPLAP immunostaining, include horizontal (h),
ganglion (rgc), displaced amacrine (da), INL amacrine (a), bipolar (b), rod (r),
cone (c) and Mller glial (m) cells. (D) Math5+ cone (arrows) and rod
(arrowheads) photoreceptors are distinguished by co-labeling with anti-hPLAP
and cone-specific PNA lectin. Non-specific labeling of pigment epithelium and
choroid reflects mouse IgG crossreactivity. (E-G) In Math5>Cre; R26floxGFP
mice, Math5+ horizontal cells (E, arrows) are marked by GFP and calbindin
immunoreactivity. The arrowhead shows a solitary Math5+ bipolar cell. (F-G)
Math5+ RGCs (arrows) and displaced amacrines (arrowheads) in the GCL are
shown in adult retinal sections (F) or P1 retinal flatmounts (G). RGCs are
distinguished by retrograde labeling of optic nerve axons with rhodamine dextran.
There is no difference in the GFP+ fraction of rhodamine dextran-labeled RGCs
between these two ages. (H) The fate of Math5>Cre- expressing progenitors in
Math5 -/- mice. hPLAP+ cells are distributed throughout the retina, but RGCs are
lacking. Vitreal hemorrhages (arrowhead) are common in Math5 -/- mice. (I) The
distribution of cell fates in the entire retina (from Jeon et al., 1998), in the Math5
lineage of wild-type mice, and in the Math5 lineage of knockout mice. The Math5
lineage is biased toward early-born cell types (RGC, horizontal, cone), although
rods are the most common fate adopted by Math5+ cells. In the Math5 knockout,
lineage-derived cells adopt all retinal fates except for RGCs. hPLAP, human
placental alkaline phosphatase; o, outer nuclear layer; i, inner nuclear layer; g,
ganglion cell layer. Scale bars, 100 m in A-B, H; 50 m in C-G.

126

127

Figure III-4. All Math5>Cre progenitors express similar levels of Cre, regardless
of cell fate. Math5>Cre lineage analysis was performed using Z/AP and
R26floxGFP reporters simultaneously, to evaluate the heterogeneity of Math5
expression among progenitors. This analysis assumes that the probability of
reporter activation in a given cell is determined by the cumulative amount of Cre
recombinase expressed by that cell. (A-B) Two models for Math5 (Cre)
expression. (A) Bimodal expression. In this model, Math5+ progenitors giving
rise to non-ganglion cell types express Cre weakly (left peak), so reporter
activation in these cells is inefficient, and consequently few of their descendants
co-express GFP and hPLAP. RGCs in the same retinas express Cre strongly
(right peak) and are expected to have high concordance. (B) Uniform
expression. In this model, every Math5+ progenitor expresses Cre strongly, so
concordance is very high for all cell types (B, right). (C) Retinas of adult
Math5>Cre; Z/AP; R26floxGFP mice immunostained for hPLAP and GFP.
Double-labeled cells (arrows) greatly outnumber single-labeled cells
(arrowheads). (D) The observed concordance between hPLAP and GFP
reporters was ~80%, which is significantly greater than expected by chance
(Cohens > 0.7). This value was similar for all cell types, indicating that Math5
is expressed at uniform levels by a subpopulation of progenitor cells, only some
of which develop as RGCs. The labeled fractions () are based on data in Table
III-1. GCL includes RGCs and displaced amacrines; INL Am, inner nuclear layer
amacrine. Scale bar, 50 m.

128

129

Figure III-5. The fate distribution of Math5+ progenitors changes over time. (A)
Cumulative BrdU labeling experiment. Math5>Cre; Z/AP embryos were
continuously exposed to BrdU from E10.5 to P0 and their retinas were collected
at P21. Nearly all Math5+ descendants (hPLAP+) are heavily labeled with BrdU,
indicating that the majority exited mitosis before P0, including lineage-labeled
cones (arrows) and rods (arrowheads). There is a distinct gradient of BrdU
labeling (birthdates) within the inner and outer nuclear layers, such that cells with
nuclei closest to the lens have earlier birthdates (brightest BrdU signal). (B)
Pulsed BrdU labeling experiment. Math5>Cre; ZAP embryos were transiently
exposed to BrdU at E15.5. Adult retinas were stained with hPLAP and BrdU
antibodies and PNA lectin. Math5+ cone (hPLAP+ PNA+ BrdU+, arrow) and
bipolar (hPLAP+ BrdU+, arrowhead) cells are indicated. (C) Cone-rod ratio plots
for birthdated hPLAP+ (red), hPLAP (blue) and combined (black) photoreceptor
groups. The ratio of cone-to-rod births decreases steadily between E14.5 and
E17.5 for hPLAP+ and hPLAP populations. The curves are parallel, indicating
that the fate of Math5+ cells changes over time, similar to other retinal
progenitors. However, the cone-to-rod ratio is 3-fold higher for Math5+
progenitors at every time point, suggesting that these cells have a fixed cone vs.
rod bias, or are shifted by 0.5 days, compared to other neurogenic cells
(hPLAP) in the same retinal environment. (D) Cone-rod ratio plot for birthdated
GFP+ (red), GFP (blue) and combined (black) photoreceptor groups in Math5 /-; Math5>Cre; R26floxGFP mice. Scale bar, 50 m.

130

131

Figure III-6. Math5 marks many of the earliest born cells in the retina. (A-C)
Window labeling analysis. (A) Embryos were exposed to pulses of EdU at E11
(onset of neurogenesis) and E11.5, and to continuous BrdU from E12 to E12.5.
Progenitors (RPCs) that continue to cycle through E12.5 are EdU+ BrdU+, while
cells that have exited mitosis between E11 and E12 are EdU+ BrdU, and
represent the earliest born cohort of retinal neurons. (B-C) Sections through the
neural retina (brackets). (B) Most early-born cells in Math5 +/- mice adopt RGC
fate (EdU+ BrdU- Brn3b+, arrows). The Brn3b- cells in this cohort are likely to
include horizontal cell precursors (arrowheads). Few Brn3b+ RGCs (arrows) are
present in Math5 -/- embryos, and the abundance of non-RGC fates increases
accordingly (arrowheads). (C) Early-born Math5-lacZ (EdU+ BrdU- gal+,
arrows) and gal (arrowheads) cells are shown in Math5 +/- (top) and Math5 -/(bottom) mice. Only ~20% of the early-born cohort expresses the Math5
transcription unit (gal+), in both genotypes. (D-E) Birthdating analysis. (D) E11
Math5>Cre; R26floxGFP embryos were exposed to a single EdU injection and
analyzed at P1. (E) Flatmounted retinas were stained for EdU and GFP and
imaged through the GCL. Strongly EdU+ cells mark the earliest born retinal
neurons. Confocal projection image (6-10 m) shows EdU+ GFP+ (arrow) and
EdU+ GFP (arrowhead) cells. Only 28% of the GCL cells born at E11 are in the
Math5+ lineage. i.p., intraperitoneal; EB, early-born. Error bars represent the
binomial standard deviation. Scale bar, 50 m.

132

133

Figure III-7. A subset of Brn3b+ RGCs derives from the Math5 lineage. (A-C)
Sections from embryonic Math5-lacZ/+ retinas co-stained for gal and Brn3b. At E11.5,
relatively few Brn3b+ cells are gal+ (A, arrow). The fraction of Brn3b+ cells expressing
Math5-lacZ (arrowheads) increases from E12.5 (B) to E13.5 (C). However, there
are many Math5-independent RGCs (arrows) at each age. (D) Histograms showing
the fraction of Math5+ cells among Brn3b+ RGCs and the fraction of Brn3b+ RGCs
within the Math5+ cohort. Error bars show the standard deviation (n = 3 sections).
The total number of cells counted at E11.5, E12.5 and E13.5 was 13, 228 and 667,
respectively. Scale bar, 50 m.

134

Figure III-8. Retrovirally marked clones exhibit symmetric and asymmetric


patterns of Math5 expression. (A) E12.5 or E13.5 retinas were explanted from
Math5 lacZ/+ embryos, flattened on polycarbonate membranes, infected at low
density with a retroviral stock to mark clonal lineages (green), and cultured for 3
days in vitro (DIV). The micrograph shows a cross-section from a representative
explant (bracket) co-stained for cytoplasmic gal (magenta) and GFP (green).
The diagram shows hypothetical 2-cell clone with gal+ cells. Each clone
reflects one informative terminal division: a symmetric [S] division which gave
rise to two Math5+ daughters (left); or an asymmetric [A] division, which gave rise
to one Math5+ and one Math5 daughter (right). (B-D) Confocal Z-stack
projections and drawings showing representative clones that are symmetric (B,
C) or asymmetric (D) with respect to Math5 expression. (E) Summary of
observed clones containing at least one Math5+ cell. Informative divisions have
a unique interpretation, and give rise to one [A] or two [S] Math5+ daughters.
Both types of divisions were identified. MIG, MSCV-IRES-GFP virus. Scale
bars: 10 m in A; 5 m in B-D.

135

136

Figure III-9. Natural history of the Math5 lineage. (A) The timing of Math5
expression shifts during retinal histogenesis. RPCs (white) shift from a
proliferative (P-P) mode of division to stem (N-P) or terminal (N-N) modes, giving
rise to neurogenic cells (gray). These express Math5 (red) either during (S,
symmetric) or after (A, asymmetric) final mitosis. During early retinal
development (<E14), Math5 is frequently expressed during G2 phase of the last
cell cycle, generating two Math5+ daughters. During later stages (>E15), Math5
is exclusively expressed by post-mitotic cells. (B) The size of the neurogenic
(birthdated) population and proportion of Math5+ cells changes during
development. At the onset of neurogenesis (E11), Math5 is expressed by
20-30% of newborn cells. The number of Math5+ cells peaks during
midgestation (E14) and rapidly diminishes (E16), while the neurogenic population
as a whole continues to expand. The temporal profile for RGC birthdates follows
similar kinetics, and reflects Math5+ and Math5 populations. (C) The fate
spectrum of Math5 lineage (red) and other neurogenic (gray) cells in wild-type
and mutant mice. The thickness and shading of arrows denotes the relative
demographic contribution of these cohorts to the mature retina.

137

138

139

1,041

1,665

horizontal

amacrine

displaced

12

2,085

467

8,800*

12,900*
< 0.1

< 0.1

28

1,648#
173,000*

11

29

31

55

(a/b) x 100

(% of cell type)

13,920*

15,570*

3,592

4,914

1,265#

(b)

total

Math5 lineage

Aa100.0

2.9

7.4

a78.5

0.9

7.0

7.9

0.5

2.2

0.6

(c) x 100

(% of retina)

cell type

2.9 (f )

< 0.002

< 0.005

0.9

0.2

0.6

0.8

0.2

0.7

0.3

(a/b) (c) x 100

(% of retina)

Math5 lineage

100

< 0.1

< 0.2

32

20

29

23

11

(a/b) (c/f ) x 100

(% of Math5 lineage)

cell type

These values are shown in Fig. 3I (middle pie chart).

Calculated from Jeon et al. (1998) and shown in Fig. 3I (left pie chart)

Among n = 8 eyes, the mean RGC labeling fraction SEM was 54 2%, with a range between 46 and 63%. The overall labeling fraction for the GCL was 40% (1167/2913 cells), which
represents 24% RGCs (700/2913 cells) and 16% displaced amacrines (467/2913 cells).

RGCs (identified by retrograde axon labeling) represent 43% of GCL neurons (1265/2913 cells). The remaining GCL cells were scored as displaced amacrines.

* Estimated using cell type ratios reported by Jeon et al. (1998) for adult C57BL/6J mice. The total number of cones counted in surveyed fields was multiplied by 35.2 to give the number
of rods, and by 3.32 for bipolar cells and 1.3 for Mller glia. The total number of GCL neurons surveyed (RGC and displaced amacrines) was multiplied by 4.78 to estimate the number
of inner nuclear layer (INL) amacrines.

RGCs, displaced amacrines, and INL amacrines were counted in 33 fields (200 X, 8 eyes, 6 mice, R26floxGFP reporter). Horizontal cells were counted in 58 sections (8 eyes, 6 mice,
R26floxGFP reporter). All other cell types were counted in 50 to 70 fields (200 X, 16 eyes, 12 mice, Z/AP reporter). Math5+ descendants detected using the Cre lineage system
comprise 2.9% of the adult retina (f ).

TOTAL

Mller glia

bipolar

rod

1,515

cone

1,198

700

RGC

INL

(a)

CELL TYPE

Math5 lineage

cells counted

Table III-1. Cell type distribution of Math5 lineage descendants in wild-type Math5 >Cre transgenic retinas

Figure III-S1. Copy number and integrity in Math5>Cre transgenes. (A)


Southern blot of EcoRI-digested DNA from nine independent transgenic lines,
hybridized with a 5' Math5 genomic probe labeled with 32P. The Math5>Cre
BAC fragment (3.5 kb) is smaller than the endogenous Math5 fragment (6.5
kb) due to the presence of an EcoRI site in the Cre cassette. Transgene copy
number was estimated from the relative intensity of these fragments, as
determined by densitometry. (B) Pulsed-field gel Southern analysis. Genomic
DNA from Math5>Cre transgenic founders 872 and 912, and purified episomal
DNA from BAC clone RP23-328P3-D1-68 were digested with NotI, separated
by PFGE, and hybridized with a Cre probe. Both founders had the predicted
84.1 kb NotI fragment, indicating that this segment is intact. Tg, transgenic.

140

Figure III-S2. Birthdating curves for cones (top graphs) and rods (bottom graphs)
in (A) wild-type Math5>Cre; Z/AP and (B) mutant Math5>Cre; R26floxGFP mice
between E14.5 and E17.5 (see Table III-S4). Single BrdU injections were given to
pregnant dams on the indicated days, and photoreceptors were analyzed in the
adult retinas. Curves show the percentage of each photoreceptor type that is
BrdU+ for the hPLAP+ or GFP+ (red), hPLAP or GFP (blue), and combined
(black) progenitor groups. The difference in scale between panels A and B
reflects the relative deficiency of late-born rods (>E17) and relative excess of
early-born cones (<E14) in Math5 -/- mice.

141

Figure III-S3. Proneural bHLH factors Neurod1 and Math5 are expressed in
overlapping subsets of progenitor cells during early retinal neurogenesis. E11.5
Math5-lacZ/+ retinal sections were immunostained for gal and Neurod1.
Approximately half of the gal+ cells co-express Neurod1 (arrows, inset). Scale bar,
50 m.

142

Table III-S1. Cell type distribution of Math5 lineage


descendants in Math5 mutant mice
Math5 lineage
cells counted

cell type
(% of Math5 lineage)

CELL TYPE

(a)

(a/b) x 100

cone

189

21

horizontal

14

amacrine

295

33

INL

187

21

displaced

108

12

358

40

rod

1.6

bipolar

17

1.9

Mller glia

21

2.3

TOTAL

894 (b)

100

All labeled cells were identified and counted in 23 fields (200X, 6 eyes,
Z/AP reporter). All cells in the GCL were assumed to be displaced
amacrine cells, as no cells in Math5 mutant retinas were positive for RGC
markers Brn3a or Brn3b.

143

144

334
257
330
257
1178

rods

cones

INL amacrines

GCL neurons*

TOTAL
1313

287

376

274

376

(a)

total hPLAP

1331

284

368

289

390

(g)

total GFP

80

82

80

84

77

d
a+gd

SEM

concordance (%)#

Standard error of the mean (SEM) calculated from concordance for n = 4 eyes. There was no significant difference among cell
types (one-way ANOVA, P > 0.15).

The incomplete concordance (<100%) in this experiment validates our assumption that Cre levels are not saturated.

There was no significant difference in the number of cells labeled by each reporter in the fields surveyed (paired t test, P > 0.1).

* RGCs and displaced amacrines.

All single and double labeled cells were counted in 18 fields (200X) representing 4 eyes from 3 Math5>Cre; Z/AP; R26floxGFP
mice. hPLAP, alkaline phosphatase; GFP, green fluorescent protein.

(d)

cell type

hPLAP+ GFP+

cells counted

Table III-S2. Dual reporter concordance for Math5>Cre labeled retinal cells

Table III-S3. Cumulative BrdU labeling experiment (E10.5 to P0)

Math5 lineage cells counted


cell type

BrdU+

total

rods

70

70

100

cones

53

53

100

INL neurons

62

65

95

GCL neurons

53

53

100

238

241

99

TOTAL

145

% BrdU+

Table III-S4. Birthdates of Math5 lineage retinal descendants


Math5 +/+ mice

cones counted (% of combined total)


time of pulse
E14.5

E15.5

E16.5

E17.5

rods counted (% of combined total)

hPLAP+

hPLAP

combined

hPLAP+

hPLAP

BrdU+

136 (12.1)

314 (27.9)

450 (40.0)

64 (0.2)

BrdU

243 (21.6)

431 (38.4)

674 (60.0)

315 (0.8)

38,700* (97.8)

39,100* (98.7)

total

379 (33.7)

684 (66.3)

1,124 (100.0)

379 (1.0)

39,200* (98.9)

39,600* (100.0)

455

(1.1)

combined
519

(1.3)

BrdU+

79 (8.3)

141 (14.5)

220 (22.7)

140 (0.4)

BrdU

214 (22.0)

537 (55.3)

751 (77.3)

317 (0.9)

32,600* (55.3)

33,000* (96.5)

total

293 (30.2)

678 (69.8)

971 (100.0)

457 (1.3)

33,700* (69.8)

34,000* (100.0)

1,073

(3.1)

1,213

(3.5)

BrdU+

28 (1.6)

100 (5.7)

128 (7.3)

383 (0.6)

BrdU

513 (29.1)

1,124 (63.6)

1,637 (92.7)

480 (0.8)

57,400* (92.4)

57,900* (93.2)

total

541 (30.7)

1,224 (69.3)

1,765 (100.0)

863 (1.4)

61,300* (98.6)

62,100* (100.0)

3,881

(6.2)

4,264

(6.8)

BrdU+

2 (0.2)

20 (1.9)

22 (2.1)

122 (0.3)

BrdU

300 (28.5)

732 (69.4)

1,032 (97.9)

264 (0.7)

34,000* (91.6)

34,300* (92.3)

total

302 (28.7)

752 (71.3)

1,054 (100.0)

386 (1.0)

36,700* (99.0)

37,100* (100.0)

2,729

(7.4)

2,851

(7.7)

For the E14.5 pulse, 16 fields (200X) were counted (3 eyes, 3 mice). For the E15.5 pulse, 14 fields (200X) were counted (4 eyes, 2
mice). For the E16.5 pulse, 24 fields (200X) were counted (5 eyes, 4 mice). For the E17.5 pulse, 15 fields (200X) were counted (4 eyes,
3 mice).
* Combined rod totals were estimated by multiplying the combined cone counts by 35.2 (Jeon et al. 1998).

Math5 -/- mice

cones counted (% of combined total)


time of pulse
E14.5

E15.5

E16.5

E17.5

GFP+

GFP

combined

rods counted (% of combined total)


GFP+

GFP

BrdU+

21 (3)

113 (17)

134 (20)

26 (0.3)

BrdU

87* (13)

455* (67)

542* (80)

83* (1)

7,699* (95)

7,782* (96)

total

108 (16)

568* (84)

676* (100)

109 (1.3)

8,025* (99)

8,134* (100)

BrdU+

8 (1)

97 (16)

105 (17)

32 (0.4)

1,066 (14)

1,098 (15)

BrdU

27* (4)

490* (79)

517* (83)

40* (0.5)

6,337* (85)

6,377* (85)

total

35 (6)

587* (94)

622* (100)

72 (1.0)

7,403* (99)

7,475* (100)

(3)

10 (1)

1,054* (95)

1,068* (96)

58* (0.4)

12,548* (94)

12,606* (94)

total

24 (2)

1,087* (98)

1,111* (100)

123 (0.9)

13,240* (99)

13,363* (100)
1,576 (18)

(4)

65 (0.5)

692 (5)

757

(4)

14* (1)

30

(4)

352

BrdU+

25 (3)

43

(4)

BrdU

(1)

33

326

combined

(6)

BrdU+

115 (1)

1,461 (16)

BrdU

73* (10)

639* (86)

712* (96)

138* (2)

7,212* (81)

7,350* (82)

total

78 (11)

664* (89)

742* (100)

253 (3)

8,673* (97)

8,926* (100)

For the E14.5 pulse, 14 fields (200X) were counted (2 mice). For the E15.5 pulse, 7 fields (200X) were counted (2 mice). For the E16.5
pulse, 9 fields (200X) were counted (2 mice). For the E17.5 pulse, 10 fields (200X) were counted (2 mice).
* Combined rod and cone totals were estimated by multiplying the measured number of inner nuclear layer nuclei by 0.326 for cones or
3.92 for rods. These ratios were determined empirically in Math5 -/- mice (data not shown).

146

CHAPTER IV
DYNAMIC EXPRESSION OF GANGLION CELL MARKERS IN RETINAL
PROGENITORS DURING THE TERMINAL CELL CYCLE

Abstract
The vertebrate neural retina contains seven major cell types, which arise
from a common multipotent progenitor pool. During neurogenesis, these cells
stop cycling, commit to a single fate, and differentiate. The mechanism and
order of these steps remain unclear. The first-born type of retinal neurons,
ganglion cells (RGCs), develop through the actions of Math5 (Atoh7), Brn3b
(Pou4f2) and Islet1 (Isl1) factors, whereas inhibitory amacrine and horizontal
precursors require Ptf1a for differentiation. We have examined the link between
these markers, and the timing of their expression during the terminal cell cycle,
by nucleoside pulse-chase analysis in the mouse retina. We show that G2 phase
lasts 1-2 hours at embryonic (E) 13.5 and E15.5 stages. Surprisingly, we found
that cells expressing Brn3b and/or Isl1 were frequently co-labeled with EdU after
a short chase (<1 hr) in early embryos (<E14), indicating that these factors,
which mark committed RGCs, can be expressed during S or G2 phases.
However, during late development (>E15), Brn3b and Isl1 were exclusively
expressed in post-mitotic cells, even as new RGCs are still generated. In
contrast, Ptf1a and amacrine marker AP2 were detected only after terminal

147

mitosis, at all developmental stages. Using a retroviral tracer in embryonic retinal


explants (E12-E13), we identified two-cell clones containing paired ganglion
cells, consistent with RGC fate commitment prior to terminal mitosis. Thus,
although cell cycle exit and fate determination are temporally correlated during
retinal neurogenesis, the order of these events varies according to
developmental stage and final cell type.

Introduction
The vertebrate neural retina is populated by seven major cell types, which
are generated in an invariant histotypic order from a common progenitor pool
(Livesey and Cepko, 2001; Wong and Rapaport, 2009). Individual cell fates are
specified by intrinsic transcriptional programs and the retinal microenvironment.
At the onset of retinal neurogenesis, on embryonic day 11 (E11) in the mouse,
the first neurons exit the cell cycle and differentiate as retinal ganglion cells
(RGCs), whose axons form the optic nerves. Mouse RGC birthdates extend from
E11 to P0 with a peak at E14 (Drager, 1985; Young, 1985a). This temporal
profile significantly overlaps the distribution of birthdates for cone photoreceptor,
horizontal and amacrine cells. Rod photoreceptors, bipolar cells and Mller glia
generally have later birthdates.
RGC development is controlled by a complex transcriptional network.
Math5 (Atoh7) is expressed transiently in multipotent precursors (Brzezinski et
al., 2012; Yang et al., 2003) and is necessary for RGC fate specification (Brown
et al., 2001; Wang et al., 2001). The Brn3b (Pou4f2) and Islet1 (Isl1)

148

homeodomain genes form two regulatory nodes that are downstream of Math5 in
the RGC differentiation hierarchy (Mu et al., 2008; Pan et al., 2008). Both factors
are thought to be expressed in newly post-mitotic cells, and are abundant in
differentiated ganglion cells (Pan et al., 2008; Pan et al., 2005; Xiang, 1998).
Brn3b appears to mark committed RGC precursors, as it is expressed exclusively
in RGCs and required for terminal differentiation (Erkman et al., 1996; Gan et al.,
1996; Qiu et al., 2008; Xiang, 1998). Isl1 is also required for RGC development
(Mu et al., 2008; Pan et al., 2008), but is expressed in a wider population of cells
(Elshatory et al., 2007a). The Brn3b+ population is essentially contained within
the Isl1 lineage (Pan et al., 2008). Amacrine and horizontal neurons are
specified in part by Ptf1a, a transiently expressed factor that acts downstream of
FoxN4 (Fujitani et al., 2006; Li et al., 2004). Brn3b and Ptf1a are likely to
assemble in opposing transcriptional complexes that regulate RGC or
amacrine/horizontal cell differentiation, respectively (Fujitani et al., 2006; Qiu et
al., 2008).
The sequential birth order of different retinal cell types reflects a shift in
the fate bias of progenitors during development (Livesey and Cepko, 2001). In
addition, the progenitor cell cycle length increases progressively (Alexiades and
Cepko, 1996; Sinitsina, 1971; Young, 1985b). Because neurons do not divide,
and differentiation occurs after the final division, it is generally thought that cell
cycle exit is strictly coupled to fate specification. However, it remains unclear
how cycle length, terminal division, and neurogenic fate are linked, and precisely
when fate commitment occurs (Dyer and Cepko, 2001b; Ohnuma and Harris,

149

2003). To explore these questions, we have determined the onset of expression


of three key regulators of RGC, horizontal and amacrine cell fate specification
(Brn3b, Isl1 and Ptf1a), relative to the terminal cell cycle. Unexpectedly, during
early development (E11.5-E13.5), we found that Brn3b and Isl1, unlike Ptf1a, can
be coexpressed prior to cell cycle exit, during late S or G2 phases. Retroviral
lineage analysis revealed multiple two-cell clones containing paired RGCs,
suggesting that these terminal progenitors represent committed ganglion cell
precursors that undergo a final mitosis. Surprisingly, the timing changes
significantly during late embryonic stages (after E15), such that Brn3b and Isl1
are expressed exclusively in post-mitotic cells. Our findings suggest that cell fate
commitment can occur before or after cell cycle exit. The order of these events
is not rigidly fixed.

Materials and Methods


Pulsed EdU/BrdU and cell cycle labeling
All mouse studies were approved by the University of Michigan Committee
on the Use and Care of Animals (UCUCA). Pregnant dams were injected with
5-bromo-2-deoxyuridine (BrdU, 100 g/g body mass) or 5-ethynyl-2-deoxyuridine
(EdU, 6.7 g/g body mass) on embryonic (E) day 11.5, 13.5, 15.5 or 16.0. BrdU
is rapidly absorbed within minutes after an intraperitoneal injection (Kriss and
Revesz, 1962). Embyros were harvested after a variable chase period, between
0.5 to 12 hours, in order to follow cells after they complete DNA synthesis (S
phase). The time of onset of a particular transcription factor expression was

150

defined by the first chase period in we detected cells positive for the particular
marker and EdU/BrdU.
Histology
Embryonic eyes were fixed for 0.5-1 hour in 2-4% paraformaldehyde
(PFA) at 22C, cryoprotected with 30% sucrose in phosphate-buffered saline
(PBS), and frozen in OCT compound (Tissue-Tek, Torrance, CA). Short fixation
was critical for immunodetection of Brn3b and cyclinD1. Cryosections (10 m)
were blocked with 10% normal donkey serum (NDS) and 1% bovine serum
albumin (BSA) in PBTx (0.1 M NaPO4 pH 7.3, 0.5% Triton X-100) for 4 hours at
22C. Sections were incubated overnight at 4C with primary antisera diluted in
PBTx with 3% NDS, 1% BSA. Sections were then washed in PBS 0.5% BSA,
incubated for 2 hours at 22C with Dylight 488- (1:1000), Dylight 549- (1:1000), or
Dylight 649- (1:500) conjugated secondary antibodies (Jackson
Immunoresearch, West Grove, PA), mounted in Prolong Gold Antifade media
(Invitrogen, Carlsbad, CA), and imaged using the Zeiss LSM510 Meta confocal
microscope at the University of Michigan Microscopy and Image Analysis Core
Facility (MIL).
The primary antibodies were: mouse anti-AP2 (1:1000, 3B5, DSHB,
Iowa City, Iowa); rat anti-BrdU (1:100, BU1/75, Harlan Seralab, Indianapolis, IN);
rabbit anti-gal (1:5000, ICN Cappel, Aurora, OH); rat anti-gal (1:500) (Saul et
al., 2008); mouse anti-Brn3a (1:50, 14A6, Santa Cruz Biotechnology, Santa
Cruz, CA); goat anti-Brn3b (1:200, sc31987, Santa Cruz); mouse anti-cyclin D1
(1:100, A-12, Santa Cruz); chicken anti-GFP (1:2000, Abcam, Cambridge, MA);
151

mouse anti-Isl1 (1:200, 39.4D5, DSHB); rabbit anti-phosphohistone H3 (1:200,


Upstate, Lake Placid, NY); rabbit anti-Ptf1a (1:800) (Fujitani et al., 2006).
Following immunostaining, EdU was detected using an azide-alkyne
cycloaddition reaction (Buck et al., 2008) (Click-iT-647, Invitrogen). For detection
of BrdU and other markers, cryosections were fully stained with primary and
secondary antibodies for the marker protein. Sections were then treated with 2.4
N HCl in PBTx for 1 hour at 22C, washed, and immunostained for BrdU.
Quantitative analysis of Brn3b onset
The percentage of EdU+Brn3b+ cells among the Brn3b+ cohort was
determined for eight different EdU chase times, at E13.5 (0.5, 1, 2, 4, 12 hrs) and
E15.5 (1, 4, 12 hrs). Double-positive cells were definitively identified by visual
examination of 3-dimensional confocal Z-stacks (225 m x 225 m x 8-12 m).
Two to six sections were analyzed for each time point, from at least 2 embryos,
representing 440 to 930 Brn3b+ cells. Error is reported as the standard deviation
among sections. The total Brn3b+ population represents both newly neurogenic
cells and existing ganglion cells. To selectively analyze neurogenic cells, we
also normalized percentages to the 12-hr chase value at E13.5. At this stage, 12
hrs is sufficient time for any EdU-labeled neurogenic cell to progress from S to
G0 phase. This single EdU pulse thus captures all newly Brn3b+ cells born within
a 5-6 hr period (length of S phase). The normalized value at 1-2 hrs reflects new
Brn3b+ cells that initiate Brn3b expression in terminal G2 phase.
Retroviral clone analysis in retinal explants

152

Retinal explant cultures and retroviral infections were performed as


described (Brzezinski et al., 2012), using established methods (Hatakeyama and
Kageyama, 2002; Wang et al., 2002b). Briefly, Math5-lacZ/+ retinas were
dissected from E12.5 or E13.5 embryos, flattened onto Nucleopore
polycarbonate membranes (0.4 m pore size, GE Healthcare, Piscataway, NJ),
and placed into Transwell culture dishes containing neurobasal media
(Invitrogen) with B27 and N2 supplements, glutamine (0.4 mM), brain-derived
neurotrophic factor (BDNF, 50 ng/mL, Peprotech, Rocky Hill, NJ), ciliary
neurotrophic factor (CNTF, 10 ng/mL, Peprotech), penicillin (50 U/mL),
streptomycin (50 g/mL), and gentamicin (0.5 g/mL).
MIG retroviral stocks were prepared in advance by calcium phosphate
transfection of the Phoenix ecotropic packaging cell line (Pear, 2001; Swift et al.,
2001) with pMSCV-IRES-GFP plasmid DNA (Van Parijs et al., 1999). Filtered
viral preparations containing polybrene (hexadimethrine bromide, 0.8 g/mL,
Sigma Aldrich, St. Louis, MO) were titered on NIH3T3 fibroblasts and diluted to
~8x105 colony-forming units per mL. One drop of retrovirus (25 L) was added
on top of each fresh explant to sparsely mark dividing cells (Roe et al., 1993) and
their descendants.
Retinal explants were cultured for 3 days at the gas-media interface at
37C under 5% CO2. After 2 days, half of the media was replaced with fresh
media. On day 3, explants were fixed for 30 minutes in 4% PFA, and processed
for cryosectioning or stained directly as wholemounts. Thick (30 m) sections, or
wholemount retinas, were immunostained for GFP and Brn3a/3b, and imaged by

153

confocal microscopy. The size and composition of clonal clusters were


determined by 3-dimensional analysis of Z-stacks. A clone was defined as a
group of directly apposed GFP+ cells that were separated by at least 4 cell
bodies from other GFP+ cells. For assessing symmetry of fate, only clones with
at least one Brn3+ RGC were scored.

Results
G2 phase is 1-2 hours long in most retinal cells
The eukaryotic cell cycle is anchored by M phase, when cytokinesis
occurs, and S phase when DNA synthesis occurs. These are separated by G1
and G2 gap phases, respectively (Fig. IV-1A) (Nurse, 2000). As a first step to
evaluate the timing of the transcription factor expression relative to cell cycle exit,
we systematically determined the onset of M phase in cycling progenitors at
different developmental stages using EdU or BrdU pulse-chase experiments. In
this experimental paradigm, cycling progenitors incorporate nucleoside analogs
EdU or BrdU during DNA replication (S phase), and these labeled progenitors
are then followed as they progress through G2 and M phases (Fig. IV-1A). To
assess the onset of M phase, we co-stained embryonic sections for
phosphohistone H3 (PH3) and EdU or BrdU to identify the shortest chase that
yielded double-labeled cells. Histone H3 is transiently phosphorylated for the
entire duration of M phase (Bradbury, 1992) No EdU+ PH3+ cells were detected
after a 1 hr EdU chase, while most PH3+ had punctate EdU labeling after a 2 hr
chase (Fig. IV-1B). These results suggest that M phase initiates at 1-2 hrs after

154

the end of S phase, in the majority of progenitors. Similar results were observed
at E15.5 (Fig. IV-1C), and are consistent with measurements of G2 length at
other embryonic times (Sinitsina, 1971; Young, 1985b).
Expression of Brn3b and Isl1, but not Ptf1a or AP2, prior to cell cycle exit
Using the pulse-chase paradigm, it is possible to unambiguously identify
the onset of expression relative to the end of S phase. Previous lineage analysis
and marker co-staining have shown that Brn3b is restricted to RGCs (Badea et
al., 2009; Qiu et al., 2008; Xiang et al., 1995), and that Ptf1a is confined to
inhibitory horizontal and amacrine cells (Fujitani et al., 2006). As a first step in
this analysis, we tested whether Brn3b and Ptf1a truly mark mutually exclusive
populations of committed precursors during early retinal development, by costaining sections from E13.5 retinas for Brn3b, Ptf1a, and the amacrine marker
AP2 (Bassett et al., 2007). We observed no overlap of Ptf1a or AP2 with
Brn3b (Fig 2A) consistent with previous results observed during late gestation
(Bassett et al., 2007; Fujitani et al., 2006; Qiu et al., 2008). As expected, many
AP2-expressing cells were contained within the Ptf1a population, and are
presumed to be precursors of inhibitory amacrine neurons.
We next determined the onset of Brn3b, Ptf1a and AP2 with respect to
the terminal cell cycle as outlined (Fig. IV-1A). At both E13.5 and E15.5, we
observed no overlap between Ptf1a or AP2 and EdU after a 4 hr chase (Fig. IV2B,C), but Ptf1a+ EdU+ cells were apparent after a 12 hr chase (Fig. IV-2D,E).
Although Ptf1a is transiently expressed and vital for development of horizontal
and amacrine cells, its actions are apparently delayed until after cell cycle exit,

155

most likely during cell migration. In contrast, at E13.5, we found that Brn3b+
EdU+ cells were present after a 1 hr chase (2-5 per high-power field, Fig. IV-3A).
All Brn3b+ cells were cyclinD1 (cycD1), including the EdU+ cohort, whereas the
vast majority of EdU+ cells (>95%) were cycD1+ (Fig. IV-3A). In most cycling
cells, cycD1 is made during G1 and down-regulated during S phase, and it
controls the G1/S transition. CycD1 is reactivated in G2 (Stacey, 2003), except
in terminally mitotic cells (Yang et al., 2006). In the developing retina, cycD1 can
be detected in all cells, except for post-mitotic neurons (G0) and a small
subpopulation of G2 cells that are likely represent the neurogenic cohort (Barton
and Levine, 2008). Accordingly, we interpret the Brn3b+ EdU+ cycD1precursors at E13.5 as neurogenic cells in late S or G2 phase of the terminal
division. Consistent with interpretation, we observed rare Brn3b+ cells at the
scleral margin (apex), in M phase (PH3+, arrows) or in adjacent pairs
(arrowheads) suggestive of a recent division (Fig. IV-3B). Furthermore, Brn3b+
EdU+ cells can be followed through the cell cycle, with symmetric Brn3b
segregation (Fig. IV-3B), and migration into the GCL with a 12 hr chase at E13.5
(Fig. IV-3C). Many of the resulting Brn3b+ EdU+ cells are in close proximity, as
presumptive daughter pairs. At E15.5, essentially no Brn3b+ EdU+ cells were
observed after 1- or 4-hr chase periods, but were readily detected throughout the
retina after a 12 hr chase (1-4 per high-power field, Fig. IV-3D). Although new
Brn3b+ cells continue to be generated throughout the retina, Brn3b is expressed
exclusively in post-mitotic cells at this stage. Quantitative analysis revealed that
only 4.8 0.9% of Brn3b+ cells at E13.5, and 1.2 0.6% at E15.5, were EdU+

156

after a 12 hr chase (Fig. IV-3E). These cells represented the sum of neurogenic
cells initiating Brn3b expression during S/G2 and G0 phase. To estimate the
fraction of Brn3b+ cells at E13.5 that initiate expression during S/G2, we
normalized the Brn3b+EdU+ fractions observed after a 0.5-4 hr chase to the 12
hr chase value (Fig. IV-3E). This analysis revealed that 50-60% of new Brn3b+
cells (accounting for mitotic doubling) initiated expression prior to cell cycle exit
(<4 hr chase).
To confirm these observations regarding the timing of RGC marker
expression relative to terminal M phase, we evaluated a second regulator of
RGC differentiation, Islet-1 (Isl1). We co-stained sections for Brn3b, Isl1 and
EdU following a 30 min chase. We observed numerous Brn3b+ Isl1+ cells that
were EdU+ at E11.5 or E13.5 (Fig. IV-4A,B), whereas no Brn3b+ Isl1+ cells were
EdU+ at E16.0 (Fig. IV-4C). These results indicate that at least two key
regulators of RGC fate are expressed prior to cell cycle exit during early (<E14),
but not during late (>E15), developmental stages.
RGC fate symmetry in marked clones
If progenitors express Brn3b and Isl1 before terminal M phase, they
should give rise to paired RGCs. To test this hypothesis, we performed a clonal
analysis. Retinal explants from E12.5 or E13.5 embryos were infected with
MSCV-IRES-GFP (MIG) retrovirus at low density to generate independent
clones. After culturing 3 days in vitro (DIV), we scored the resulting GFP+ clones
for their size and number of RGCs, using Brn3a or 3b immunoreactivity and
cellular morphology to identify ganglion cells (Fig. IV-5A). The Brn3a paralog is

157

functionally interchangeable with Brn3b at the protein level (Pan et al., 2005) and
its spatiotemporal expression pattern is highly overlapping with Brn3b and Brn3c
(Xiang et al., 1995; Xiang et al., 1993).
In 12 explanted retinas, we observed approximately 250 GFP+ clones,
which were 1-16 cells in size. Fourteen (~5%) included at least one Brn3+ RGC.
Among these informative clones, we observed two-cell clones with paired (Fig.
IV-5B,C) or single (Fig. IV-5D) RGCs. Of the 12 neurogenic divisions that could
unambiguously be scored among 14 clones, six were symmetric, resulting in
paired RGCs (Fig. IV-5E). In principle, paired RGCs could arise if both daughter
cells independently adopt the RGC fate, stochastically or in response to the local
retinal environment. Given the very small number of GFP+ ganglion cells
observed in these explants (n = 24), and the high frequency of symmetric fating (6
of 12 divisions), this pattern is unlikely to reflect independent events paired by
chance (2 > 40, df = 2, P < 0.001 for a Poisson distribution). Instead, this
statistical clustering, and the expression of definitive RGC markers in G2 and
through cytokinesis (Fig. IV-3), strongly argue that many RGCs arise in pairs and
descend from progenitors that commit to the ganglion fate during their final cell
cycle, prior to terminal mitosis.

Discussion
The kinetics of the retinal progenitor cell cycle have been extensively
characterized in rodents and other vertebrate species by window and cumulative
nucleoside labeling, cell counting, and percent labeled mitosis (PLM) methods

158

(Alexiades and Cepko, 1996; Fujita, 1962; Li et al., 2000; Sinitsina, 1971; Young,
1985b). In addition, a variety of nuclear factors responsible for generating
histotypic diversity in the retina have been identified through loss- and gain-offunction genetic analysis, and expression studies (Ohsawa and Kageyama,
2008; Swaroop et al., 2010). However, the basic question of when the cell fate
decision is made relative to the terminal cell cycle has remained largely
unanswered. In this paper, we have integrated expression and cell cycle
analysis to determine the precise time that transcription factors controlling RGC,
amacrine and horizontal fates appear during the last (neurogenic) cell cycle.
RGC fate commitment can occur before cell cycle exit
At early developmental times (<E14), we detected Isl1 and Brn3b in many
cells during late S or G2 phases of the terminal cell cycle (Figs. 3,4), similar to
Math5-lacZ (Brzezinski et al., 2012), and in contrast to previous reports (Pan et
al., 2008; Xiang et al., 1993). This interpretation is based on three convergent
results: the precise timing of PH3 (Fig. IV-1), Brn3b and Isl1 (Fig. IV-4) onset
relative to the end of S phase, the mutually exclusive pattern of cycD1 and Brn3b
expression (Fig. IV-3A), and the presence of Brn3b+ M phase cells (Fig. IV-3B).
There are three possible explanations for the observation of RGC markers
in dividing cells. First, the immunopositive cells may reflect low-level leaky
transcription of Brn3b or Isl1 that is not biologically meaningful. This seems
unlikely given that the expression in S/G2 phase progenitors is often as strong as
that observed in post-mitotic cells (e.g. arrows in Fig 4A,B). Second, Isl1 and
Brn3b may not accurately mark committed RGC precursors. Indeed, Isl1 is also

159

expressed in differentiating amacrine and bipolar precursors, and has a role in


the development of both of these interneuron classes (Elshatory et al., 2007a;
Elshatory et al., 2007b). In contrast, Brn3b is not co-expressed with Ptf1a or
AP2, which mark nascent horizontal and amacrine interneurons, and it is made
in most, but not all RGCs (Fig. IV-2A). Complementary co-expression and
lineage analyses, using Cre recombinase or stable histogenic reporters, have
also demonstrated that Brn3b is made only in RGCs (Badea et al., 2009; Bassett
et al., 2007; Fujitani et al., 2006; Qiu et al., 2008). In the absence of Brn3b,
progenitors adopt non-RGC fates (Badea et al., 2009; Qiu et al., 2008).
Overexpression of Brn3b in progenitors promotes RGC fate and suppresses
amacrine differentiation (Feng et al., 2011; Liu et al., 2000; Qiu et al., 2008).
Taken together, these results strongly suggest that Brn3b identifies committed
RGCs, and does not label multipotent progenitors. Third, the RGC fate decision
may precede terminal mitosis, and occur during G2. According to this
interpretation, Brn3b+ Isl1+ EdU+ cells represent progenitors committed to
develop as RGCs. In general, progenitors may undergo symmetric or
asymmetric divisions to generate one or two post-mitotic daughters, respectively
(Huttner and Kosodo, 2005). Our pulse-chase (Fig. IV-3) and clonal analyses
(Fig. IV-5) suggest that many divisions giving rise to RGCs are symmetric and
terminal. This observation is somewhat surprising considering that the retina is
rapidly expanding during the period of early development. However, G2 fate
commitment may be necessary for prompt deployment of RGCs, which establish
a pioneering scaffold for axon pathfinding (Pittman et al., 2008; Raper and

160

Mason, 2010). Alternatively, the relatively early onset of Brn3b expression may
reflect the position of RGCs as the first sampled fate in the histogenetic
sequence (Wong and Rapaport, 2009).
Our clonal data are generally consistent with previous observations of
terminal divisions in the early retina. In rodent and frog lineage analysis, many
two-cell clones have been observed during early development, but typically these
cells have discordant fates (Holt et al., 1988; Turner et al., 1990; Wong and
Rapaport, 2009). Notably, among 114 two-cell clones analyzed in these reports,
only one contained paired RGCs (Holt et al., 1988). During early zebrafish
development, most ath5-expressing progenitors, monitored by time-lapse
imaging of mosaic ath5-GFP embryos, undergo a terminal division and give rise
to at least one ganglion cell (Poggi et al., 2005). The daughters typically have
discordant fates, but paired RGCs may arise from wild-type cells in an ath5
mutant environment.
Similar to our finding of mitotic Brn3b+ cells (Fig. IV-3B), RA4+ cells were
observed in terminal mitotic anaphase in the chick retina (Waid and McLoon,
1995), and migrating in pairs shortly after mitosis (McLoon and Barnes, 1989).
Although originally interpreted as RGC precursors, it is no longer clear that the
RA4 neurofilament antigen marks ganglion cells exclusively (Gutierrez et al.,
2011). Beyond these observations, there is precedent for G2 commitment during
neurogenesis of other cell types. In the chick retina, committed horizontal cell
precursors typically arrest in G2 and undergo a final non-apical mitosis (Boije et
al., 2009), which may explain the presence of paired horizontal cell clones of a

161

single subtype (Rompani and Cepko, 2008). In the ferret cerebral cortex,
progenitors lose their competence to respond to environmental signals following
terminal G2 phase (McConnell and Kaznowski, 1991). Similarly, heterochronic
analysis of dissociated rat retinal cells suggests that progenitors commit to the
amacrine fate (VC1.1+) during G2/M of the last cell cycle (Belliveau and Cepko,
1999).
Post-mitotic fate plasticity in the retina
In this study, the amacrine marker Ptf1a was only detected in post-mitotic
cells, at both E13.5 and E15.5 (Fig. IV-2B-E). In Ptf1a mutant mice carrying Cre
knock-in alleles, many lineage-marked cells expressed Brn3b and adopted RGC
fates (Fujitani et al., 2006). Taken together, these results suggest a degree of
post-mitotic plasticity, such that some amacrine cells solidify their identity after
the terminal division. Likewise, photoreceptor and bipolar precursors retain the
ability to switch fates after cell cycle exit (Adler and Hatlee, 1989; Brzezinski et
al., 2010; Ng et al., 2011; Oh et al., 2007). Indeed, the bZIP factor NRL, which
instructs rod photoreceptor fate, initiates expression soon after terminal M phase,
throughout development (Akimoto et al., 2006); data not shown).
We observed a gradual restriction in the expression of Brn3b and Isl1, as
well as Math5-lacZ (Brzezinski et al., 2012), to post-mitotic cells during late
gestation (Fig. IV-3D). At these ages (>E15), many RGCs are still generated,
including migrating EdU+ Brn3b+ cells (Drager, 1985; Young, 1985a). In the
early (<E14) retina, we observed both symmetric and asymmetric patterns of
RGC marker expression among daughter cells in clones, suggesting that some

162

cells commit to the RGC fate after terminal mitosis in the early retina (Fig. IV-5E)
(Brzezinski et al., 2012). The difference in the timing of RGC marker onset
between the early and late retina cannot be explained by a change in the length
of G2, which is relatively static (Fig. IV-1), or by the progressive lengthening of
the cell cycle during development (Alexiades and Cepko, 1996; Sinitsina, 1971;
Young, 1985b). Indeed, if transcription initiates after G1 or S phase with a fixed
time delay, then the onset of expression following terminal S phase should be
advanced (closer to S), rather than delayed as we observed during the course of
development. It is more likely that progenitor fate decisions occur before or after
cell cycle exit, depending on the lineage or developmental stage.
What could explain these shifts in expression kinetics? Environmental
signals elaborated by differentiating neurons, such as Delta ligand or Shh,
increase as RGCs accumulate (Ahmad et al., 1997; Wang et al., 2005; Yang,
2004; Yu et al., 2006) and could delay the onset of RGC transcription factor
expression. Alternatively, the onset may be intrinsically programmed to shift over
time, and is potentially regulated by the same upstream cascades that control
cell cycle exit. Further studies are needed to distinguish these mechanisms.
Nonetheless, it is clear that cell fate determination and cell cycle exit, though
correlated, are not strictly causally related events.
Acknowledgements
The authors are grateful to Dellaney Rudolph and Melinda Nagy for technical
support; to Helena Edlund for Ptf1a antisera; to Chris Edwards, and the UM
microscopy and image analysis laboratory staff, for technical advice; to Sean

163

Morrison for the MIG retroviral construct; to Joe Brzezinski, Nadean Brown, Chris
Chou, Sally Camper and David Turner for valuable discussions and critical
reading of the manuscript. This research was funded by National Institutes of
Health (NIH) R01 grant EY14259 (TG). LP was supported by NIH T32 grants
EY13934 and GM07863.

164

Figure IV-1. Timing of cell cycle progression in the mouse retinal neuroepithelium
at E13.5 and E15.5. (A) EdU pulse-chase experiments. Phosphohistone H3 (PH3)
is expressed during M phase, while cyclin D1 is primarily expressed during G1 and
early S phases. After an EdU pulse, labeled S phase cells progress into G2, M
and G1/G0 phases. (B) Diagram of interkinetic nuclear migration, showing the
positions of progenitor nuclei at each stage of the cell cycle. Following terminal M
phase, G0 daughters segregate vertically according to their cell fate. Ganglion
cells (rgc) migrate toward the base (vitread), while photoreceptors (pr) migrate
toward the apex (sclerad). (C-D) E13.5 and E15.5 embryos stained for PH3 and
EdU or BrdU following a 1-2 hour chase. No PH3+ cells are EdU+ after a 1 hour
chase, but most are EdU+ or BrdU+ after a 2 hour chase. Therefore, M phase
initiates at 1-2 hours after S phase in most cells, at both of developmental stages.
nbl, neuroblastic layer; gcl, ganglion cell layer. Scale bar, 50 m.

165

Figure IV-2. Co-expression and onset analysis of amacrine and horizontal


markers Ptf1a and AP2. (A) Developing horizontal or amacrine neurons do not
express Brn3b. Sections from E13.5 retinas coimmunostained for Brn3b, AP2
and Ptf1a. Although many Brn3b+ cells are present, none co-label with the
horizontal and inhibitory amacrine marker Ptf1a, or the pan-amacrine marker
AP2, suggesting that Brn3b specifically marks RGCs during early development.
(B-E) Sections from E13.5 (B,D) or E15.5 (C,E) embryos co-stained with Ptf1a,
AP2 and EdU following a 4 hr (B,C) or 12 hr (D,E) chase. No Ptf1a+ or AP2+
cells are EdU+ after a 4 hr chase, suggesting that these factors are expressed
exclusively in post-mitotic (G0) cells. Many Ptf1a+ EdU+ cells are apparent after
a 12 hr chase at both E13.5 and E15.5 (arrows, insets), indicating that this factor
initiates expression in migratory cells. Many AP2+ cells are Ptf1a+
(arrowheads), marking the inhibitory amacrine population. Scale bar, 50 m.

166

167

Figure IV-3. The onset of Brn3b and Isl1 expression within individual cells is
progressively delayed during retinal development. (A) Sections from E13.5
embryos co-stained for Brn3b, cyclinD1 (cycD1) and EdU following a 1 hr chase.
In some cells, Brn3b is co-localized with EdU (arrows), but not with the G1/earlyS phase marker cycD1, indicating that Brn3b expression initiates in late S or G2
phase. Insets show two cells that are Brn3b+ EdU+ cyclinD1. One has a
punctate EdU staining pattern, indicating incorporation in late-replicating DNA
regions at the end of S phase. (B) Sections from E13.5 embryos stained for
Brn3b and the mitotic marker PH3 or the nuclear counterstain DAPI. Brn3b+
cells in M phase (arrows) and presumptive daughter pairs (arrowheads) are
indicated. (C-D) Sections from E13.5 or E15.5 embryos co-stained for Brn3b and
EdU following the indicated chase period. At E13.5 (C), many newly generated
Brn3b+ cells are co-labeled with EdU after a 2, 4 or 12 hr chase. By 12 hrs,
some EdU-labeled cells migrate to the nascent GCL, and many Brn3b+ EdU+
cells are in close proximity as presumptive daughter pairs. At E15.5 (D), no
Brn3b+ EdU+ cells are observed after 1 hr chase, and very few are detected
after a 4 hr chase, but Brn3b+ EdU+ are readily detected after a 12 hr chase
(arrows). (E) Quantitative analysis of Brn3b and EdU co-labeling. The
abundance of co-labeled cells is plotted as a percentage of all Brn3b+ cells (left
scale) or neurogenic Brn3b+ cells (right scale, normalized to the 12 hr chase
value at E13.5). Approximately 30% of newly Brn3b+ cells initiated expression
prior to cell cycle exit at E13.5 (gray line). In contrast, few Brn3b+ cells are
co-labeled with EdU prior to 12 hrs at E15.5, indicating that RGCs are specified
later at this developmental stage. Scale bar, 50 m.

168

169

Figure IV-4. Co-expression of Brn3b and Isl1 during or shortly after the
terminal cell cycle. (A-C) Sections from E11.5, E13.5, and E16.0
embryos co-stained for Brn3b, Isl1 and EdU following a 30 min chase.
At E11.5 (A) and E13.5 (B), multiple Brn3b+ Isl1+ EdU+ cells are
observed (arrows), indicating that RGC markers can initiate expression
prior to cell cycle exit, in late S or G2 phase. In contrast, at E16.0 (C), all
Brn3b+ Isl1+ cells are EdU (examples marked by arrowheads in D-F). The
insets show orthogonal Z-stack views of magnified Brn3b+ Isl1+ EdU+ cells.
Scale bar, 50 m.

170

Figure IV-5. Paired ganglion cells can be generated from retinal progenitors by
symmetric terminal division. (A) Retinas were explanted from E12.5 or E13.5
embryos, infected at clonal density with MSCV-IRES-GFP (MIG) retrovirus
(green), and cultured for 3 days in vitro (DIV). The micrograph shows a crosssection from a representative explant (bracket) coimmunostained for Brn3a
(magenta) and GFP (green). The schema shows informative two-cell GFP+
clones: a symmetric [S] clone containing two Brn3+ RGCs, and an asymmetric
[A] clone with one Brn3+ RGC. (B-D) Confocal Z-stack projections and drawings
from representative symmetric (B, C) or asymmetric (D) clones containing RGCs.
Most Brn3+ cells have long processes (arrows), confirming that they are
differentiated RGCs. (E) Summary of observed clones containing at least one
Brn3+ cell. In these experiments, RGCs were identified using Brn3a (12 clones)
or Brn3b (2 clones) antisera interchangeably. Among 12 informative divisions,
an equal number with symmetric and asymmetric fates was observed. Scale
bars, 10 m in A; 5 m in B-D.

171

172

CHAPTER V

PUSHING THE ENVELOPE OF RETINAL GANGLION CELL GENESIS:


CONTEXT DEPENDENT FUNCTION OF MATH5 (ATOH7)

Abstract
The basic-helix-loop helix factor Math5 (Atoh7) is required for retinal
ganglion cell (RGC) development. However, only 10% of Math5-expressing cells
adopt the RGC fate, and most become photoreceptors. In principle, Math5 may
actively bias progenitors towards RGC fate or passively confer competence to
respond to instructive factors. To distinguish these mechanisms, we
misexpressed Math5 in a wide population of precursors using a Crx BAC or 2.4
kb promoter, and followed cell fates with Cre recombinase. In mice, the Crx
cone-rod homeobox gene and Math5 are expressed shortly after cell cycle exit,
in temporally distinct, but overlapping populations of neurogenic cells that give
rise to 85% and 3% of the adult retina, respectively. The Crx > Math5 transgenes
did not stimulate RGC fate or alter the timing of RGC births. Likewise, retroviral
Math5 overexpression in retinal explants did not bias progenitors towards the
RGC fate or induce cell cycle exit. The Crx>Math5 transgene did reduce the
abundance of early-born (E15.5) photoreceptors two-fold, suggesting a limited
cell fate shift. Nonetheless, retinal histology was grossly normal, despite
widespread persistent Math5 expression. In an RGC-deficient (Math5 knockout)

173

environment, Crx>Math5 partially rescued RGC and optic nerve development,


but the temporal envelope of RGC births was not extended. The number of
early-born RGCs (before E13) remained very low, and this was correlated with
axon pathfinding defects and cell death. Together, these results suggest that
Math5 is not sufficient to stimulate RGC fate. Our findings highlight the robust
homeostatic mechanisms, and role of pioneering neurons in RGC development.
Introduction
The vertebrate retina is a highly ordered structure composed of six major
types of neurons and one type of glia. These originate from a common
progenitor pool (Holt et al., 1988; Turner and Cepko, 1987; Turner et al., 1990)
and include retinal ganglion cells (RGCs), rod and cone photoreceptors,
amacrine, horizontal and bipolar interneurons, and Mller glia. At the onset of
retinal neurogenesis, embryonic day 11 (E11) in mice, multipotent retinal
progenitor cells (RPCs) begin to exit the cell cycle and differentiate.
Birthdating studies, in which nucleoside analogs are used to identify
progenitors exiting the cell cycle, have defined a fixed, but overlapping, order for
the generation of these major cell classes in vertebrates (Rapaport et al., 2004;
Sidman, 1961; Young, 1985a). RGCs are the first to exit the cell cycle, at E11 in
mice, with peak birthdates at E14 and termination by P0 (Drager, 1985). This
early temporal profile overlaps significantly with those of cone, horizontal, and
amacrine neurons. Rods, Mller glia and bipolar cells have characteristically
later birthdates. In mice, the distribution of rod births peaks in the neonatal
period, but the tails of the distribution extend across most of the histogenic

174

period, from E12.5 to P10 (Carter-Dawson and LaVail, 1979; Swaroop et al.,
2010), because rods compose ~80% of the mature retina (Jeon et al., 1998).
Heterochronic co-culture and transplantation experiments, in which early
embryonic and late RPCs were cultured in unequal ratios, have suggested that
fate determination is largely a cell intrinsic process (Belliveau and Cepko, 1999;
Rapaport et al., 2001; Reh, 1992; Watanabe and Raff, 1990). Indeed,
progenitors cultured at low density can develop into each of the major cell
classes, with similar diversity and proportions as the intact retina, in the absence
of environmental feedback signals (Adler and Hatlee, 1989; Cayouette et al.,
2003; Reh and Kljavin, 1989). However, extrinsic signals can influence
progenitor cell cycle dynamics and override fate decisions in vivo (Cepko, 1999;
Ezzeddine et al., 1997; Kim et al., 2005; Yang, 2004). Collectively, these
observations are consistent with a temporal, or serial, competence model for
retinal development (Cepko et al., 1996; Livesey and Cepko, 2001; Reh and
Cagan, 1994; Wong and Rapaport, 2009). According to this model, RPCs pass
through discrete competence states over time, in which they can adopt a limited
number of cell fates. Within each state, the decision to exit the cell cycle and the
final histotypic choice are influenced by extrinsic signals.
Two prototypical intrinsic factors important for development of mouse
RPCs into specific types of neurons are the cone-rod homeodomain (HD) factor
Crx and the basic-helix-loop-helix (bHLH) factor Math5 (atonal homolog Atoh7).
Crx, and closely related factor Otx2, are expressed in during or shortly after the
terminal cell cycle in tripotential precursors that give rise to photoreceptors and

175

bipolar cells (Furukawa et al., 1997; Muranishi et al., 2011). The precise role of
Crx remains unclear. In mice, Crx expression initiates at E12.5 and is necessary
for proper development of photoreceptors, and may be partially redundant with
Otx2 in conferring competence for photoreceptor specification (Chen et al., 1997;
Furukawa et al., 1999; Nishida et al., 2003; Sato et al., 2007). Crx works in
concert with other transcription factor to regulate photoreceptor gene expression
(reviewed in (Hennig et al., 2008; Swaroop et al., 2010)), and Crx is abundant in
adult rods, cones and bipolar cells. The 5 regulatory DNA for Crx has been
extensively characterized, and a critical 2.4 kb promoter region is thought to
faithfully recapitulate the endogenous Crx pattern. This segment has been used
to drive Cre, lacZ, and regulators of rod photoreceptor specification in transgenic
mice (Cheng et al., 2006; Furukawa et al., 2002; Koike et al., 2005; Nishida et al.,
2003; Oh et al., 2007).
Like Crx, the role of Math5 in fate specification has not been fully
elucidated. Math5 (Atoh7) is a single-exon gene that is transcribed by retinal
progenitors in a spatiotemporal pattern that mirrors RGC births (Brown et al.,
1998; Brzezinski et al., 2012; Prasov et al., 2010). This factor is transiently
expressed by RPCs during or after their terminal cell cycle (Brzezinski et al.,
2012; Feng et al., 2010; Kiyama et al., 2011) and is required for RGC
development. Math5 mutant mice have very few RGCs (<5%) and lack optic
nerves (Brown et al., 2001; Wang et al., 2001). Apart from the deficiency of
RGCs, all other retinal cell classes are preserved (Brown et al., 2001; Brzezinski
et al., 2005). The mRNA profiles of Math5 mutant retinas are altered, with

176

downregulation of genes associated with RGC differentiation (Mu et al., 2005).


Lineage tracing experiments have established that Math5-expressing cells
contribute to 3% of the adult retina, and every major cell class (Brzezinski et al.,
2012; Feng et al., 2010; Yang et al., 2003). Together, these data suggest Math5
acts as an essential competence factor for RGC development. Orthologous
genes in zebrafish, chicken, and frog have similar functions, lineage properties,
and expression patterns (Kanekar et al., 1997; Kay et al., 2001; Liu et al., 2001;
Matter-Sadzinski et al., 2001; Poggi et al., 2005), and mutations in human
ATOH7 have been linked to optic nerve aplasia and retinal vascular disease
(Ghiasvand et al., 2011; Khan et al., 2011; Prasov et al., 2012).
Despite these expression and phenotypic analyses, the precise role of
Math5 in RGC development remains unclear. Gain-of-function studies in frog
and chick, and mouse embryonic stem cells, have yielded mixed results. In
these systems, Math5 biases proliferating progenitors towards RGC fates when
over-expressed during early developmental stages (Brown et al., 1998; Kanekar
et al., 1997; Liu et al., 2001; Moore et al., 2002; Yao et al., 2007), but promotes
other cell fates when expressed during late development, or in a cross-species
context (Brown et al., 1998; Moore et al., 2002). In general, the interpretation of
these experiments is confounded by the tendency of proneural bHLH factors to
drive cell cycle exit when overexpressed (Farah et al., 2000).
To circumvent these limitations and critically assess the role of Math5 in
biasing RGC development, we generated transgenic mice that ectopically
express Math5 in a large number of retinal progenitors and newly post-mitotic

177

neurons, under control of a mouse Crx promoter fragment (Crx>Math5 Tg) or


bacterial artificial chromosome (Crx>Math5 BAC), with a bicistronic Cre lineage
tracer. Although endogenous Crx and Math5 genes mark overlapping
populations, and appear to be co-expressed in some cells, profound
overexpression of transgenic Math5 did not stimulate RGC production or alter the
profile of RGC births. Instead, the number of early-born photoreceptors was
reduced. Despite sustained high-level expression of Math5 in photoreceptors
and bipolar cells, retinal histology and cell type distribution were grossly normal.
Likewise, no substantial RGC bias was observed in retinal explants infected with
a Math5-expressing retrovirus. In mutant mice, the Crx>Math5 transgenes
rescued RGC development. However, because endogenous Crx expression
initiates somewhat later that Math5, early-born RGCs were scarce, and some
rescued ganglion cells exhibited pathfinding defects or apoptosis during
development. These results suggest that Math5 action is context dependent.
Our findings also warrant a re-examination of previous results obtained using
conventional Crx transgenes.

Materials and Methods


Conventional and Bacterial Artifical Chromosome (BAC) transgenes
To ectopically express Math5 in a wide population of retinal cells, we
generated a conventional Crx>Math5-IRES-Cre bicistronic transgene. We
assembled mouse Math5 cDNA and Cre recombinase coding sequences,
separated by an internal ribosomal entry site (IRES2) and followed by a SV40

178

polyA signal. The mouse Crx promoter and proximal regulatory region were
amplified by PCR and inserted upstream as a 2.4 kb XhoI-SalI fragment
(Furukawa et al., 2002; Oh et al., 2007). A matched Crx>Cre transgene was
then generated from the Crx>Math5-IRES-Cre plasmid by precise deletion of
Math5 and IRES sequences using the single-strand oligonucleotide (ss oligo)
recombineering method (Thomason et al., 2007), with a 70 nt antisense oligo
(Suppl. Table 1) and AscI selection.
To faithfully express Math5 in the endogenous Crx pattern, we generated
BAC transgenes by RED recombineering (Lee et al., 2001). The targeting
construct was assembled, with short (400 bp) 5 and 3 homology arms (H)
flanking a Math5-IRES-Cre-FRT-amp-FRT cassette. This was equivalent to the
conventional transgene, but included an FRT-amp-FRT selection cassette (Gene
Bridges, Heidelberg) downstream of the SV40 polyA signal. The 5 homology
arm extends from Crx intron 1 to the exon 2 initiation (ATG) codon, while the 3
homology arm contains sequence from Crx intron 2. A matched control (CreFRT-amp-FRT) was then generated by ss oligo recombineering, with the 70 nt
antisense oligo (Suppl Table 1) and AscI selection.
Linearized targeting plasmids were used in parallel to target mouse BAC
clone RP23-81H17 by RED-mediated homologous recombination in strain
SW105 (Warming et al., 2005) after heat induction. This 219 kb BAC contains
134 kb 5 and 69 kb 3 DNA flanking the Crx gene. Targeted BAC clones were
selected on ampicillin and chloramphenicol plates at 30C, and verified by
junctional PCR and DNA sequencing. The amp seletion cassette was then

179

deleted by arabinose induction of Flpe recombinase, leaving a solitary FRT site


(Andrews et al., 1985). Homogeneity and integrity of the resulting clones was
verified by ampicillin sensitivity, junctional PCRs, restriction mapping, and
pulsed-field gel electrophoresis.
Purified circular DNA from BAC transgene constructs or linearized plasmid
DNA from conventional constructs was injected into fertilized (C57BL/6J SJL/2)
F2 or R26floxGFP (JAX stock 004077 reporter strain, (Mao et al., 2001)) x
B6SJLF1/J oocytes by the UM Transgenic Animal Core. Founders were
identified by transgene-specific PCR genotyping (Suppl. Table 1), and lines were
maintained by crossing to C57BL/6J or R26floxGFP reporter strains. We
analyzed 2 founders and 2 lines for each conventional transgene (Crx>Cre Tg
and Crx>Math5 Tg), and 3 lines for each BAC transgene. The most extensively
characterized transgenes in this report were Crx>Math5 Tg 251, Crx>Cre Tg 352
control, Crx>Math5 BAC 60, and Crx>Cre BAC 764 control.
RNA analysis
Duplex and competitive triplex RT-PCRs were performed as described
(Prasov et al., 2010) to compare the levels of transgene-derived and endogenous
mRNAs. Total RNA was extracted from embryonic eyes (E14.5) or adult (P21)
tissues of transgenic or wild-type animals using Trizol reagent (Invitrogen,
Carlsbad, CA). First-strand cDNA was generated by high-fidelity reverse
transcription (RT, Transcriptor, Roche) at 50C and used as template for PCR,
with primers and conditions in Suppl. Table 1. Triplex competitive RT-PCRs
used a common 6-carboxyfluorescein (FAM)-labeled forward primer in Crx exon
180

1, and two reverse primers (Suppl. Table 1). Dual products were closely
matched for size and G+C content, and analyzed using a 3730XL capillary
electrophoresis unit (Applied Biosystems, Carlsbad, CA) and Gene Marker
software (SoftGenetics, State College, PA).
Quantitative RT-PCRs were performed using custom Taqman probes and
Universal Taqman Mastermix (Applied Biosystems), and were analyzed on the
ABI 7600 Real Time PCR System. Critical cycle threshold levels were
normalized to Gapdh internal controls, using two-fluorophore (VIC and FAM)
detection. Fold activity was calculated using the ddCt method (Livak and
Schmittgen, 2001) and reported relative to the Crx>Math5 BAC expression level.
Expression copy-number levels were calculated for conventional and BAC
transgenes by normalizing to endogenous Crx values, using the mean ratio
determined in triplex competitive RT-PCRs (k). Measurements were obtained
using independent RNA pools from 2-5 mice of each genotype.
Histology
For section immunostaining, eyes or embryonic heads were fixed in 2-4%
paraformaldehyde (PFA) 0.1 M NaPO4 pH 7.3 for 30-60 min at 22C, processed
through a 10-30% graded sucrose series, embedded in OCT (Tissue-Tek,
Torrence, CA) and cryosectioned at 10 m. For flatmount preparations, eyes of
P1 or adult mice were removed and fixed in 4% PFA for 5 min. The optic nerves
were then transected, and the retinas were teased apart from other ocular
tissues, fixed in 4% PFA for 25 min. After immunostaining, retinas were incised

181

with 6-8 radial cuts and flattened with the ganglion cell layer (GCL) facing
upward.
For immunodetection, slides or whole retinas were blocked in a solution of
10% normal donkey serum (NDS), 1% bovine serum albumin (BSA) in PBTx (0.1
M NaPO4 pH 7.3 0.5% Triton X-100) for 1-4 hrs. To reduce mouse-on-mouse
background associated with mouse monoclonal primary antibodies, donkey antimouse IgG Fab fragments were added at 0.8 mg/mL to some blocking reactions.
Primary antibodies were applied overnight at 4C and diluted in 3% NDS 1% BSA
in PBTx. Sections or retinas were then washed in PBS, incubated for 2 hrs at
22C with Dylight-conjugated secondary antibodies and 4',6-diamidino-2phenylindole (DAPI), and mounted in Prolong Gold Antifade (Invitrogen, Grand
Island, NY). Slides were imaged using the Zeiss LSM510 Meta confocal system
or an Olympus BX-51 epifluorescence microscope.
The primary antibodies were mouse anti-AP2 (1:1000, DSHB, Iowa City,
IA); rabbit anti-gal (1:5000, ICN Cappel, Aurora, OH); rat anti--galactosidase
(1:500, (Saul et al., 2008)); rat anti-BrdU (BU1/75, 1:100, Harlan Seralab,
Indianapolis, IN); mouse anti-calbindin (CB-955, 1:500, Sigma, St. Louis, MO);
rabbit anti-cleaved-caspase 3 (1:100, Cell Signaling, Beverly, MA); sheep antiChx10 (1:250, Exalpha, Shirley, MA); mouse anti-Cre (clone 7.23, 1:300,
Covance, Princeton, NJ); rabbit anti-Crx (1:1000, (Zhu and Craft, 2000));
chicken anti-GFP (1:2000, Abcam, Cambridge, MA); mouse anti-hPLAP
(monoclonal 8B6, 1:250, Sigma); mouse anti-PKC (MC5, 1:100, Sigma); rabbit
anti-mCar (1:500, Millipore, Billerica, MA); mouse anti-syntaxin (HPC-1, 1:1000,

182

Sigma); rabbit anti-M-opsin (1:1000, Millipore); rabbit anti-S-opsin (1:5000,


(Applebury et al., 2000)); rabbit anti-phosphohistone H3 (1:400, Upstate, Lake
Placid, NY); rabbit anti-rhodamine (1:500, Invitrogen); rabbit anti-Sox9 (1:250,
Millipore); rabbit anti-TuJ1 (MRB-435P, 1:2000, Covance). The Crx antibody
appears to cross-react weakly with Otx2 antigen (Brzezinski et al., 2010), most
likely through a shared LDYKDQ sequence in the Crx 14-residue peptide
immunogen (Zhu and Craft, 2000).
For detection of BrdU (5-bromo-2-deoxyuridine) and other antigens,
cryosections were fully stained with primary and secondary antibodies to the
other markers. Sections were then treated with 2.4 N HCl in PBTx for 1hr at
22C, and immunostained for BrdU. Likewise, EdU (5-ethynyl-2-deoxyuridine)
was detected after immunostaining, using an azide-alkyne cycloaddition reaction
(Buck et al., 2008) and with Click-iT-647 reagents (Invitrogen).
For fine histology, mice were perfused transcardially with 2% PFA and
1.25% glutaraldehyde. The eyes were removed, post-fixed overnight at 22C,
dehydrated, embedded in glycol methacrylate plastic resin (JB-4, Polysciences,
Warrington, PA), sectioned at 4 m with a Leitz 1512 rotary microtome, and
stained with basic fuchsin and methylene blue. Paraffin or cryosections (5-10
m) of eyes or optic nerves were stained with hematoxylin and eosin as
described (Brown et al., 2001).
Retrograde axon labeling of RGCs
RGCs were definitively marked by retrograde axon labeling with
rhodamine dextran (Brzezinski et al., 2012; Rachel et al., 2002). Eyes from adult
183

or P1 mice were removed and immersed in Hanks balanced salt solution


containing calcium, magnesium and 1 mM glucose (HBSSG). Optic nerves were
transected within 1 mm of the sclera, and lysine-fixable tetramethyl rhodamine
dextran 3,000 MW powder (Molecular Probes, Eugene, OR) was applied directly
to the cut site. The eyes were positioned with severed optic nerves facing
downward against cubes of surgifoam (Ethicon, Somerville, NJ) saturated with
3% L--lysophosphatidyl choline (LPC, Sigma) and rhodamine dextran. These
were sealed with 1 % agarose, and incubated en bloc in aerated HBSSG for 1
hour at 22C. The surgifoam was then removed, and the eyes were incubated
overnight in HBSSG under the same conditions. Rhodamine-labeled eyes were
fixed in 4% PFA for 4 hrs at 22C and processed for sectioning or stained as
whole retina preparations. In some experiments, the signal was enhanced by
indirect immunofluorescence staining with anti-rhodamine antibody.
Cre lineage and dual reporter concordance analysis
To trace the descendants of cells expressing Cre recombinase, transgenic
mice were crossed to R26floxGFP or Z/AP (JAX stock 003919, (Lobe et al.,
1999)) reporter strains, which activate cytoplasmic GFP (green fluorescent
protein) and membrane-tethered hPLAP (human placental alkaline
phosphatase), respectively, after excision of floxed upstream stop signals.
Retinal sections or flatmounts were co-stained for histotypic antigen markers,
hPLAP and/or GFP. Cell types were identified by characteristic laminar position,
morphology, and marker co-localization. RGCs were clearly distinguished from
displaced amacrines by retrograde axon labeling. To assess the heterogeneity in

184

the level of Crx transgene (Cre) expression among progenitors giving rise to
different cell types, we conducted dual reporter concordance experiments, as
described (Brzezinski et al., 2012). Coexpression of GFP and hPLAP was
scored in the outer nuclear (ONL), inner nuclear (INL) and ganglion cell (GCL)
layers of adult triple transgenic mice (Crx>Math5 BAC or Crx>Cre BAC; Z/AP;
R26floxGFP).
Quantitative assessment of RGCs in transgenic mice
Retinal ganglion cells were counted in Crx>Math5 Tg and non-transgenic
littermates at P0 and P22, using Brn3a or retrograde axon labeling to mark
RGCs. The fractional contribution of RGCs to the GCL (DAPI nuclei) was
determined from 20 sections (400X) representing n = 2 eyes of each P22
genotype, and 12-22 sections (200X) representing n = 4-6 eyes for each P0
genotype.
To evaluate transgenic rescue of RGC development in mutants,
Crx>Math5 Tg or Crx>Math5 BAC mice were crossed to Math5 knockout (KO)
mice (Atoh7tm1Gla, (Brown et al., 2001)) for two or more generations. Eyes from
informative embryonic, neonatal and adult littermates were immunostained as
sections or flatmounts. Retinal cell death was assessed at E16.5 using activated
Caspase-3 staining (Gown and Willingham, 2002). RGCs and apoptotic cells
were counted in 18 sections (200X) representing n = 6 eyes of each genotype.
EdU pulse-chase and birthdating analysis

185

To evaluate the overlap between Crx+ and Math5+ cell populations, and
compare the timing of Crx and Math5 expression, pregnant dams carrying E13.5
or E15.5 Math5-lacZ/+ embryos were given to a single intraperitoneal injection of
EdU (6.7 g/g body mass). Embryos were harvested after a 4-hr chase and their
retinas were stained for Crx, gal and EdU. For other short-term labeling
experiments, a single pulse of EdU or BrdU was given to pregnant dams 1 hr
prior to harvest.
To assess alterations in the fate distribution of neurogenic cells exiting
mitosis on different days, pregnant dams carrying Crx>Math5 Tg (line 251) and
non-transgenic control embryos were given a single injection of BrdU (100 g/g
body mass) on E12.5, E13.5 or E15.5. Retinal sections from the resulting mice
were stained for BrdU at P21, and the distribution of strongly BrdU+ cells among
GCL, INL and ONL layers was determined. For the E15.5 pulse, we counted 24
sections (200X) from n = 6 eyes of each genotype, representing a total of 816
birthdated cells in Crx>Math5 Tg mice and 1005 birthdated cells in control mice.
To evaluate late-stage RGC births, dams carrying Crx>Math5 BAC (line 60) and
control embryos were pulsed with EdU at E17.5 and harvested at P22. Likewise,
Crx>Math5 Tg (line 251) pups with Math5 KO and heterozygous (het) genotypes,
and non-transgenic littermates, were pulsed with BrdU at P1 and harvested at
P22.
RGC birthdating curves were generated for rescued and control
littermates by giving single EdU pulses at E11 and E12, E13.5, E15.5, or E17.5.
The resulting pups (four genotypes) were harvested at P1 and their retinas were

186

stained for Brn3a and EdU as flatmounts. Two 0.05 mm2 areas in the central
retina were imaged as confocal Z-stacks through the GCL for each flatmount
preparation. The density of RGCs (Brn3a+ cells per mm2) and birthdated RGCs
(Brn3a+ EdU+ cells per mm2) was determined by direct counting. The
normalized RGC birth fraction was calculated by dividing the number of RGCs
born at each time point by the sum of RGCs born in all four time points (E11-E12,
E13.5, E15.5, E17.5).
Statistics
Comparisons were made using a two-tailed Students t-test in Microsoft
Excel in cases where equal variance was observed. The Welch t-test was used
for comparisons among groups of unequal variance. Errors are reported for
biological replicates as SDM (standard deviation of the mean) unless otherwise
noted. Jitter plots were generated with Prism software (Graphpad, La Jolla, CA).

Clonal analysis in retinal explants


Retinal explant cultures and retroviral infections were performed as
described (Brzezinski et al., 2012) using standard methods (Hatakeyama and
Kageyama, 2002; Wang et al., 2002b). Briefly, retinas were dissected from
E13.5 wild-type embryos, flattened onto Nucleopore polycarbonate membranes
(0.4 m pore size, GE Healthcare, Piscataway, NJ) and transferred to Transwell
culture dishes containing neurobasal media (Invitrogen) with B27 and N2
supplements, glutamine (0.4 mM), BDNF (50 ng/mL, Peprotech, Rocky Hill, NJ),

187

CNTF (10 ng/mL, Peprotech), penicillin (50 U/mL), streptomycin (50 g/mL), and
gentamicin (0.5 g/mL).
The Math5-IRES-GFP retroviral plasmid was constructed by inserting a
Math5 cassette in the bicistronic MSCV-IRES-GFP (MIG) retroviral vector (Van
Parijs et al., 1999). MSCV-IRES-dnMAMLGFP was generated by replacing the
GFP cassette in MIG with dnMAMLGFP. This encodes a fusion protein with
residues 12-72 of mouse MAML1 (mastermind-like) at the N-terminus and GFP
at the C-terminus (Maillard et al., 2004). Retroviral stocks were prepared in
parallel by calcium phosphate transfection of the Phoenix ecotropic packaging
cell line (Pear, 2001; Swift et al., 2001) with plasmid vectors. Polybrene
(hexadimethrine bromide, 0.8 g/mL, Sigma Aldrich, St. Louis, MO) was added to
filtered media containing infectious particles. These viral stocks were titered on
NIH3T3 cells and diluted to ~8 105 CFU (colony forming units) per mL.
Transductions were performed by pipetting one drop (~25 L) on top of each
fresh explant, to sparsely mark dividing cells (Roe et al., 1993) and their
descendants.
Infected explants were cultured for 7 days at the gas-media interface at
37C under 5% CO2. Half of the media was replaced with fresh media on days 2,
4 and 6. After one week in culture, explants were fixed in 4% PFA for 30 min and
processed for cryosectioning. Serial thick (30 m) cryosections were
immunostained for GFP and Brn3a, and imaged as 3-dimensional confocal Zstacks. Clones were scored for size (number of GFP+ cells) and composition
(number of Brn3a+ RGCs). A clone was defined as an isolated group of directy

188

apposed GFP+ cells, separated by at least 4 cell bodies from other GFP+ cells.
We scored 3-4 explants per virus, giving a total number of 70 (IRES-GFP), 60
(Math5-IRES-GFP) and 52 (IRES-dnMAML) clones, respectively.

Results
Crx and Math5 are expressed in comparable neurogenic cell populations
Crx and Math5 progenitor cell populations give rise to 85% (rods, cones
and bipolar cells) and 3% of the mouse adult retina (Brzezinski et al., 2012;
Furukawa et al., 1997; Jeon et al., 1998), respectively (Fig. V-1A). To compare
these factors directly, we examined their overlap and onset of expression relative
to the terminal S phase, using a lacZ allele (Atoh7tm1Gla, (Brown et al., 2001)) as a
proxy for Math5. We found that Crx and gal are expressed in distinct, but
overlapping, cohorts of cells at both E13.5 and E15.5 (Fig. V-1B,C), consistent
with the lineage profile of Math5 descendants, over half of which are
photoreceptors (Brzezinski et al., 2012). Using EdU or BrdU pulse-chase
analysis at these time points, we determined that a small number of Crx and gal
double-positive cells at E13.5 and E15.5 were also labeled with EdU after a 4-hr
chase (Fig. V-1). At these developmental stages, four hours is sufficient time for
some cells progress through S, G2, and M phases and enter G0 (Prasov and
Glaser, 2012; Sinitsina, 1971; Young, 1985b). Given that both of these factors
are expressed during or after the terminal division (Brzezinski et al., 2012;
Muranishi et al., 2011); and data not shown), these results suggest that Crx and
Math5 can be made in the same cells, simultaneously or sequentially, at the time

189

when cell fate is determined. While Math5 is short-lived, Crx expression persists
in a broad population of differentiated photoreceptor and bipolar cells (Brzezinski
et al., 2012; Furukawa et al., 2002; Furukawa et al., 1997).
Crx>Math5 conventional and BAC overexpression systems
To critically assess the role of Math5 in biasing progenitor cell fate, we
generated transgenic mice with ectopic Math5 expression. Given the overlap of
endogenous Crx and Math5 expression, and the similar nature of these
neurogenic cells, we chose the Crx promoter to broadly express Math5 in a large
population of early post-mitotic precursors. The 2.4 kb Crx promoter has been
extensively characterized and is thought to drive specific expression in
photoreceptor precursors, and mature rods, cones and bipolar cells (Furukawa et
al., 2002). However, to limit position effects and ensure faithful Crx expression,
we also built a BAC transgene containing 134 kb regulatory DNA upstream of the
Crx start site (exon 1), and 69 kb downstream of the polyadenylation signal (exon
3). We then generated conventional and BAC transgenic mice, termed
Crx>Math5 Tg and Crx>Math5 BAC, respectively (Fig. V-2A). The transgenes
express Math5 and Cre from bicistronic transcripts. The Cre recombinase
allowed us to trace the fate of cells expressing transgene-derived Math5 using
the R26floxGFP reporter (Mao et al., 2001), which makes GFP after excision of a
loxP-flanked stop signal. In parallel, we generated matched Crx>Cre and
Crx>Cre BAC control transgenic mice to confirm the Crx lineage, and isolate
Math5 effects.

190

Multiple independent insertions ( 4) of each transgene were analyzed for


GFP and Cre expression, after crossing to R26floxGFP reporter mice (Fig. V-2B).
These patterns were consistent, indicating that position effects were minimal in
the vast majority of the transgenic lines (Fig. V-S1 and data not shown). For
subsequent analysis, we chose a single representative line for each of
transgene.
To evaluate transgene fidelity, we co-immunostained sections from adult
and embryonic mice for GFP, Cre, and Crx antigens (Fig. V-2B). In both
Crx>Math5 Tg mice (line 251) and Crx>Cre Tg mice (line 352), we observed
cumulative expression (GFP) throughout the neural retina, RPE (retinal
pigmented epithelium), and ciliary body, but not in lens or scleral tissue (Fig. V2B, Fig. V-S2). GFP expression was also evident in the pineal gland (Fig. V-S2),
and in embryonic forebrain regions (data not shown). Within the retina, all cell
layers, and the vast majority of cells, were labeled with GFP by both conventional
transgenes. In contrast, the patterns of Cre and Crx immunostaining were much
more restricted. Cre antigen was completely co-extensive with Crx in adult
photoreceptors and bipolar cells, although the relative expression levels varied
among INL cells. Likewise, in the embryonic retina, all Cre+ cells were Crx+ (Fig.
V-2B and data not shown). These results indicate that the Crx 2.4 kb promoter
drives strong expression in photoreceptor and bipolar precursors, but is also
active in multipotent progenitors or in other post-mitotic cells. Together, these
patterns are most consistent with high-level expression in Crx domains (Chen et
al., 1997; Furukawa et al., 1997), and leaky expression in the domains of closely

191

related homeodomain factor Otx2 (Bovolenta et al., 1997; Simeone et al., 1993).
In Crx>Cre BAC mice (line 60) and Crx>Math5 BAC mice (line 764), cumulative
transgene expression (GFP) was confined to the retina (Fig. V-2B), and was not
observed in the RPE or ciliary body as noted with the conventional transgenes
(Fig. V-S2). Again, Cre+ cells were co-extensive with the Crx population of cells
at both adult and embryonic time points (Fig. V-2B). Within the adult retina, only
a few scattered cells were labeled with GFP in the inner INL and GCL by the
Crx>Cre BAC and Crx>Math5 BAC transgenes, as expected. Overall, the
patterns of BAC and conventional transgene expression were very different,
suggesting that the Crx 2.4 kb promoter is active beyond the endogenous Crx
domain.
To further characterize the BAC and conventional transgenes, we
assessed the distribution and level of Math5 mRNA expression by RT-PCR. In
the Crx>Math5 Tg, transcripts were evaluated in various tissues using primers
specific to transgenic (Tg) or endogenous Math5 (Fig. V-2B). Both species were
confined largely to the eye, with a low level of endogenous Math5 mRNA in the
brain (Saul et al., 2008). Only transgenic Math5 was detected in the adult retina,
consistent with the known patterns of Math5 and Crx expression (Brown et al.,
1998; Brzezinski et al., 2012; Chen et al., 1997; Furukawa et al., 1997). To
quantitatively compare the levels of BAC and conventional Crx>Math5 transgene
expression, we first determined the ratio of Crx>Math5 BAC and endogenous Crx
transcripts using a triplex competitive RT-PCR assay (Prasov et al., 2010) with a
common end-labeled forward primer located in Crx exon 1 (Fig. V-S3). We found

192

that BAC-derived Math5 transcripts were present at 53 3% (k) the level of


endogenous Crx transcripts in the adult retina, or approximately single-copy
expression levels. We next measured Math5 (a) and Crx (b) RNAs relative to
Gadph in Crx>Math5 BAC, Crx>Math5 Tg and control adult retinas by TaqMan
quantitative PCR. We found that Math5 levels in Crx>Math5 Tg retinas were 16fold higher than those in BAC transgenic retinas. As expected, Math5 was not
detected in controls and the level of Crx mRNA did not vary between genotypes.
In the adult Crx>Math5 Tg retinas, the level of Math5 expression was thus 8-fold
higher than endogenous Crx (a k/b). The differences observed between BAC
and conventional transgenic mice (Fig. V-2B) may thus reflect differences in the
pattern and/or level of expression.
Lineage analysis of Crx>Math5 transgenes
To determine whether ectopic Math5 biases progenitors towards particular
cell fates, we evaluated the distribution of cell types in Crx>Math5 BAC and
Crx>Math5 Tg retinas compared to control Crx>Cre BAC and Crx>Cre Tg
retinas, by examining GFP staining pattern of double transgenic mice carrying
the R26floxGFP reporter. As expected, each transgene labeled all
photoreceptors and bipolar cells. Persistent Math5 expression did not grossly
alter the subtype distribution of cones or bipolar cells (Fig. V-S4).
In addition to rods, cones and bipolar cells, the Crx>Math5 BAC and
Crx>Cre BAC control transgenes marked a small number of ganglion, horizontal
and amacrine neurons, with frequency that varied from region to region in the
retina and among cell types (Fig. V-3). Early cell types, particularly horizontal

193

neurons, were rarely marked by Crx>Cre BAC or Crx>Math5 BAC transgenes


(Fig. V-3E,F), consistent with the onset of Crx expression at E12.5 (Chen et al.,
1997; Furukawa et al., 1997). Likewise, in conventional Crx>Math5 Tg and
Crx>Cre Tg mice, the vast majority of retinal cells were marked with GFP,
including all major cell types. As noted with the BAC transgenes, fewer RGCs
and horizontal cells were marked compared to other cell types (Fig. V-S5 and
data not shown).
In principle, the labeled RGCs, horizontal and amacrine cells in BAC
transgenic mice could represent one of three classes: [1] lineal descendents of
proliferating progenitors that expressed low or high levels of the transgene, [2]
rare Crx+ precursors that adopted non-photoreceptor or bipolar fates, [3] cells
whose fate was shifted due to the action of Math5. To distinguish these
mechanisms, we used two approaches. First, we assessed the heterogeneity of
Cre expression levels among cells using a dual-reporter concordance paradigm
(Brzezinski et al., 2012). Adult triple transgenic Crx>Math5 BAC or Crx>Cre BAC
mice, carrying R26floxGFP and Z/AP (Lobe et al., 1999) reporters, were
immunostained for GFP and hPLAP (Fig. V-S6). Among photoreceptors in the
ONL and bipolar cells in the outer INL, concordance for GFP and hPLAP was
uniformly high for each transgene (nearly 100%). However, among neurons in
the inner INL and GCL, concordance was low (<40%), indicating that these cell
types expressed a low level of Cre, and thus recombined stochastically at only
one reporter locus. Second, we analyzed the cycle kinetics and spatial
distribution of informative cells in E15.5 embryos carrying Crx>Math5 BAC and

194

R26floxGFP. After a 1 hr EdU pulse, no Cre+ EdU+ cells were detected (Fig. VS7A). However, many EdU+ cells were observed in GFP+ vertical stripes (Fig.
V-S7B), where most labeled Brn3b+ RGCs were localized (Fig. V-S7C). These
findings are consistent with the general clustering of GFP+ cells in the adult
retinas, and suggest a clonal origin (Reese et al., 1999; Turner and Cepko,
1987). Similar patterns were observed for the Crx>Math5 Tg embryos, but the
relative abundance of GFP+ progenitors was a much greater (data not shown).
Crx>Math5 expression does not stimulate RGC genesis
Given the limitations of using the Cre lineage reporter to assess cell fate,
we employed other metrics to assess the effects of broad Math5 overexpression
on the overall fate distribution. We focused primarily on ganglion cells for these
experiments, because Math5 is necessary for RGC development, and we used
Crx>Math5 Tg, because this transgene is expressed at much higher levels than
the BAC counterpart (Fig. V-2D). We observed a similar abundance of Brn3a+ or
rhodamine dextran-labeled RGCs in Crx>Math5 Tg and control mice throughout
development (Fig. V-4A-C). At P0, prior to the neonatal culling of RGCs (Erkman
et al., 2000; Farah and Easter, 2005; Galli-Resta and Ensini, 1996; Young,
1984), 56 5% SD of GCL neurons in Crx>Math5 Tg retinas were RGCs
(Brn3a+), similar to controls (59 5% SD, P = 0.4, Fig. V-4D). Likewise, at P22,
no significant difference was observed in the fraction of RGCs (rhodamine
dextran-labeled) in the GCL between the two genotypes (41.1 0.1% SD for wildtype, 43 8% SD for Crx>Math5 Tg, P = 0.8 Fig. V-4D) or previous reports
(41 4% SD (Jeon et al., 1998)). We conclude that broad ectopic Math5

195

expression does not promote RGC fate, alter the survival of RGCs, or drive Brn3
expression in a Math5 wild-type retina.
Crx>Math5 expression alters the distribution of early born cell types
Given the persistent high level of expression of the Crx>Math5 Tg in rods,
cones and bipolar cells, we assessed overall retinal histology in plastic sections
(Fig. V-4E). The morphological features of photoreceptor nuclei, inner and outer
segments, and other layers were not affected at this level. However, subtle fate
shifts might occur during development, which are counterbalanced by
homeostatic feedback mechanisms. To assess these effects, we used a
birthdating approach. Embryos were exposed to single BrdU pulses at E12.5,
E13.5, or E15.5 and their retinas were analyzed at P22. At each time point,
fewer birthdated nuclei were observed in the ONL of Crx>Math5 Tg mice than
controls (Fig. V-5A, Suppl. Fig 8). For quantitative analysis, we focused on the
E15.5 time point, as these litters contained a sufficient number of animals of each
genotype for statistical comparisons. We counted the number of strongly BrdU+
nuclei in ONL, INL and GCL layers. We observed a 2-fold decrease in the
fraction of birthdated photoreceptors (ONL cells) in Crx>Math5 Tg retinas
(21 3% SD) compared to controls (43 2% SD, t-test P < 10-3), with a
corresponding increase in the INL and to lesser extent GCL (Fig. V-5A).
In principle, the loss of early-born photoreceptors could be due to cell
death from persistent Crx>Math5 Tg expression or a small, bona fide fate shift.
To distinguish these mechanisms, we evaluated Crx>Math5 Tg and control
retinas for apoptosis by activated Caspase-3 immunostaining. At multiple time

196

points between E13.5 and P0, we observed 1-2 apoptotic cells per field (200X) in
both Crx>Math5 Tg and control retinas (Fig. V-S9), consistent with previous
studies of cell death in the embryonic retina (Vecino et al., 2004). Crx>Math5 Tg
is thus unlikely to induce cell death in photoreceptor precursors. Instead, ectopic
Math5 appears to shift the fates of some early rod and cone photoreceptors.
Crx>Math5 expression does not extend the temporal profile of RGC births
In Crx>Math5 Tg mice, the expression of Math5 is extended through
postnatal development, whereas endogenous Math5 mRNA is downregulated by
P0 (Brown et al., 2001; Brzezinski et al., 2012). To test whether prolonged
expression in neurogenic cells extends the profile of RGC births, we pulsed
Crx>Math5 Tg mice with BrdU at P1 and harvested eyes at P22. In the central
two-thirds of the retina, no BrdU+ cells were detected in the GCL of either
genotype (Fig. V-5B), consistent with the completion of displaced amacrine and
RGC genesis in these areas by P1 (Farah and Easter, 2005; LaVail et al., 1991;
Reese and Colello, 1992; Voinescu et al., 2009; Young, 1985a). Similarly, in
flatmounts of adult Crx>Math5 BAC retinas that were exposed to an EdU pulse at
E17.5, we observed very few, if any,EdU+ RGCs (Fig. V-5C). Therefore, the
envelope of RGCs births is not extended by prolonged Math5 expression.

Retroviral Math5 does not induce RGC fate or cell cycle exit in retinal
explants
We also tested whether Math5 can bias proliferating progenitors towards
RGC fate, using a retroviral vector to transduce cultured retinal explants. E13.5

197

retinas were infected at low density ex vivo with MIG vectors. The resulting
single-copy proviruses express Math5 and GFP, or GFP alone from the potent
MSCV LTR promoter (Hawley, 1994) (Fig. V-6A). After 7 days in culture, we
counted the number of Brn3a+ RGCs (Fig. V-6B-C). Among 70 clones infected
with the IRES-GFP retrovirus, 4/272 GFP+ cells were Brn3a+ (1.5 0.7%
binomial SD, Fig. V-6D), in accord the fraction of RGCs produced by in vivo
clonal analysis (Turner et al., 1990). Among 60 clones infected with Math5IRES-GFP retrovirus, 6/262 GFP+ cells were Brn3a+, which is not different from
those transduced with GFP alone (2.2 0.9% binomial SD, Fishers exact
P = 0.34). Because overexpression of bHLH factors can promote cell cycle exit
and differentiation (Farah et al., 2000), we also evaluated clone size. We found
that the size distribution did not vary significantly between explants infected with
these viruses (Fig. V-6E, 2 test P = 0.6 for df = 4), indicating that high-level
single-copy expression of Math5 does not significantly promote cell cycle exit.
As a positive control, we also analyzed explants infected with an MSCV-IRESdnMAMLGFP retrovirus. The dnMAMLGFP fusion protein autonomously blocks
Notch signaling by interfering with the NICD-CSL transcriptional complex
(Maillard et al., 2004). Inhibition of Notch activity is associated with premature
cell cycle exit and stimulation of RGC fate among early progenitors (Austin et al.,
1995; Nelson et al., 2007). Among 52 clones, we detected a modest increase in
the fraction of Brn3a+ RGCs (Fig. V-6D, 3/60 GFP+ cells, 5 3%, P = 0.11). We
also observed a significant reduction in clone size, with all cells deriving from
one- or two-cell clones (Fig. V-6E, 2 test P < 10-8 for df = 4). These results

198

confirm that blockade of Notch signaling by dnMAMLGFP drives cell cycle exit,
and that our explant system is sufficiently robust to detect this effect. Ectopic
Math5 expression in progenitors does not stimulate cell cycle exit or significantly
promote RGC fate in this system, consistent with our transgenic overexpression
findings (Fig. V-4).
Crx>Math5 expression partially rescues the RGC deficiency in Math5 KO
mice
Although Crx>Math5 Tg does not stimulate RGC genesis in the wild-type
environment, cells expressing ectopic Math5 may be prevented from adopting
the RGC fate by strong negative feedback from nascent RGCs (Austin et al.,
1995; Belliveau and Cepko, 1999; Waid and McLoon, 1998; Wang et al., 2005;
Zhang and Yang, 2001). Indeed, we have observed that many Brn3a+ are
generated in E13.5 Math5 KO retinal explants transfected with human ATOH7
(Prasov et al., 2012). We therefore crossed Crx>Math5 Tg and Crx>Math5 BAC
transgenes onto the Math5 KO background to examine the potential of Crxdriven Math5 to stimulate RGC genesis, heterochronically and heterotopically
(Fig. V-1A), in a deficient environment.
We evaluated retinal flatmounts for the density of mature RGC cell bodies,
axons and fascicles (Fig. V-7A-C). In Math5 heterozygous mice, the vast
majority of axons made radial projections to the optic disc and were fasciculated
(Fig. V-7A,B), and the GCL contained many Brn3a+ RGCs (Fig. V-7C). In
contrast, Math5 KO retinas had vastly reduced axon density, in accord with
previous estimates (Lin et al., 2004). The residual axons were largely

199

unfasciculated and exhibited pathfinding defects similar to those in Brn3b -/- mice
(Badea et al., 2009; Gan et al., 1999) (Fig. V-7B and Fig. V-S10). Among 10
flatmounts examined from Math5 KO eyes, no Brn3a+ cells were observed in any
area (Fig. V-7C). The Crx>Math5 Tg transgene, and to a lesser extent the
Crx>Math5 BAC, were variably capable of rescuing axons and preventing
fasciculation defects in Math5 KO retinas. However, rescue was less
pronounced with successive generations of mice, ostensibly due to epigenetic
reductions in transgene expression levels (Garrick et al., 1998). The Crx>Math5
Tg, but not the Crx>Math5 BAC, transgene was capable of restoring Brn3a
expression among some adult ganglion cells. As expected, all rescued Brn3a+
ganglion cells were derived from progenitors that expressed the Crx>Math5 Tg
transgene (GFP+), whereas only ~40% of Brn3a+ cells were GFP+ in the wildtype (Fig. V-7D). Finally, we observed small optic nerves in rescued mice
carrying the Crx>Math5 Tg (Fig. V-7E), but not in Math5 KO controls. Together,
these results suggest that the Crx>Math5 Tg transgene, which expresses high
levels of Math5 in early post-mitotic precursors and low levels in progenitors, can
partially rescue RGC genesis and axonal guidance defects in Math5 KO mice.
Crx>Math5 expression alters the RGCs birth profile in Math5 KO mice
Given the differences in the timing of Crx and endogenous Math5
expression, we tested whether the rescued RGCs were born within the same
temporal envelope as native RGCs, which are only generated prenatally. We
first exposed P1 pups to a single pulse of BrdU, and stained their retinas at P21.

200

Few, if any, GCL cells were birthdated at P1 in Math5 KO mice carrying


Crx>Math5 Tg (Fig. V-7F).
To fully explore the temporal profile of RGC births and the extent of RGC
genesis in rescued mice, we generated partial birthdating curves (Young, 1985a)
for Crx>Math5 Tg; Math5 KO mice and littermate controls (Fig. V-8). Crx>Math5
Tg; Math5 KO mice were crossed to Math5 heterozygotes and the pregnant
dams were given two injections of EdU at E11 and E12 to mark the earliest born
RGCs (Fig. V-8A), or a single injection of EdU at E13.5, E15.5, or E17.5 (Fig. V8B-D). Retinas were harvested from the resulting pups and stained as
flatmounts for Brn3a and EdU at P1, a time point before the neonatal RGC
culling period (Farah and Easter, 2005), but after all RGCs have been generated
(Figs. 5B and 7F). In Math5 KO and Crx>Math5 Tg; Math5 KO mice, very few
Brn3a+ RGCs were born during early development (Fig. V-8A). In all 4
genotypes, the majority of RGCs were born between E13.5 and E15.5 (Fig. V8B,C), consistent with RGC birthdating curves for wild-type retinas (Drager,
1985; Farah and Easter, 2005; Young, 1985a). Furthermore, no E17.5
birthdated Brn3a+ cells were detected in any of these mice, after careful
examination of confocal Z-stacks through the GCL (Fig. V-8D).
The birthdating data and RGC counts reveal four important trends (Fig. V8E,F). First, the curves for Math5 heterozygous (het) and Crx>Math5 Tg; Math5
het mice were nearly identical, suggesting that Crx>Math5 Tg does not alter the
profile of RGC births (2 test P = 0.26 for df = 2). Second, the normalized
distribution of RGC births for Crx>Math5 Tg; Math5 KO and Math5 het mice

201

differed significantly (2 test P = 0.01 for df = 2). In particular, very few RGCs
were generated in the rescued mice (Crx>Math5 Tg; Math5 KO) at the earliest
developmental times, and a larger number of RGCs were born at E15.5. Third,
RGCs were generated within the same temporal envelope in all 4 genotypes,
suggesting that this time window is fixed. It does not depend on Math5, and
cannot be shifted by protracted or elevated Math5 expression.
Fourth, the absolute number of rescued RGCs at P1 in Crx>Math5 Tg;
Math5 KO mice was increased 2.3-fold in comparison to control Math5 KO mice
(Welch t-test P = 0.02, Fig. V-8G), but this effect was variable, and the number of
RGCs was low in comparison to Math5 hetetozygotes (12%, 1160 250 Brn3a+
cells per mm2, P < 10-8). The Crx>Math5 transgene did not increase the RGC
density in Math5 hets (10,400 400 Brn3a+ cells per mm2 vs. 10,800 600,
Students t-test P = 0.54), consistent with the results above (Fig. V-4). Likewise,
the number of RGCs was significantly reduced in control Math5 KO mice
compared to heterozygotes (5%, 500 50 Brn3a+ cells per mm2, P < 10-8), as
previously reported (Lin et al., 2004).
Increased cell death and optic nerve defects in rescued Crx>Math5 Tg mice
The variability and incomplete rescue of RGCs in Crx>Math5 Tg mice on
the Math5 KO background was surprising. To explore the mechanism underlying
this variability, we analyzed optic nerve development during early
embryogenesis, using TuJ1 to identify RGC axons. At E15.5 and E17, we
observed coalescence of TuJ1+ fibers and formation of optic nerves in
Crx>Math5 Tg; Math5 KO, but not in Math5 KO retinas (Fig. V-9A,B). These

202

rescued optic nerves were much thinner than wild-type controls. In some cases,
these nerves exhibited severe pathfinding defects as they exited the retina, and
formed large knot structures (Fig. V-9A, arrowhead). Because RGCs that fail to
properly establish connections in the CNS are eliminated (O'Leary et al., 1986),
we evaluated cell death by activated Caspase-3 immunostaining at E16.5, near
the end of RGC genesis. We observed a significant increase in cell death in
Math5 KO mice compared heterozygous controls (Fig 9C-D, 9 1 vs. 1.1 0.2
Casp3+ cells per field, Welch t-test P < 0.001), consistent with previous reports
(Feng et al., 2010). In Crx>Math5 Tg; Math5 KO mice, there was a further
increase in cell death compared to Math5 KO retinas (12.9 0.6 Casp3+ cells per
field, P = 0.016). The vast majority of dying cells were located in the GCL (Fig. V9C), suggesting that aberrant RGCs are eliminated. We also observed knots of
RGC axons in Crx>Math5 BAC; Math5 KO mice, consistent with the small degree
of rescue by this transgene (Fig. V-9E).

Discussion
The patterns of Crx>Cre BAC and Crx 2.4 kb transgene expression
The Crx 2.4 kb promoter has been used to drive Cre, lacZ, Nrl or Nr2e3
expression (Cheng et al., 2006; Furukawa et al., 2002; Koike et al., 2005; Nishida
et al., 2003; Oh et al., 2007), and Crx BACs have been used to drive GFP,
hPLAP or lacZ expression (Muranishi et al., 2010; Samson et al., 2009). From
these studies, it is clear that both types of transgenes are expressed at high
levels in photoreceptors and bipolar cells, and their precursors. However,

203

thorough lineage data reflecting cumulative transgene expression are lacking. In


this study, as control for Math5 overexpression, we characterized the cumulative
expression of Crx 2.4 kb promoter and Crx>Cre BAC transgenes using
R26floxGFP and Z/AP reporters. At the level of Cre immunodetection (Fig. V-2),
both transgene formats recapitulated endogenous Crx expression throughout
development, and were restricted to known Crx domains in the eye and pineal
gland (Fig. V-2, Fig. V-S2) (Chen et al., 1997; Furukawa et al., 1997). However,
when Cre activity was assessed using highly sensitive reporters, the Crx 2.4 kb
promoter and, to a much lesser extent the Crx>Cre BAC, were found to mark all
major cell types in the retina, contrary to previous reports (Koike et al., 2005;
Nishida et al., 2003). Multiple independent insertions gave similar results (Fig. VS1 and data not shown), so the ectopic patterns cannot be attributed to
chromatin position effects.
There were notable differences between conventional and BAC
transgenes. First, the GFP reporter was detected in the RPE, ciliary body and
embryonic brain with conventional, but not with BAC transgenes (Fig. V-S2 and
data not shown), and GFP was more broadly expressed within the neural retina.
These differences are likely to reflect leaky expression of the Crx 2.4 kb promoter
in Otx2 domains. Otx2 and Crx are evolutionary paralogs (Plouhinec et al.,
2003), and Otx2 is expressed in the RPE, ciliary body, retinal progenitors, and
embryonic forebrain (Bovolenta et al., 1997; Brzezinski et al., 2010; Muranishi et
al., 2011; Simeone et al., 1993). Given their structural and functional similarity,
common evolutionary origin, and overlapping expression in photoreceptors and

204

the pineal, at least some transcription factors are likely to regulate both genes,
and bind within this 2.4 kb sequence. Indeed, the zebrafish Crx ortholog is
expressed in proliferating retinal progenitors (Shen and Raymond, 2004), and it
has been suggested that bovine Crx, along with Otx2, regulate gene expression
in the RPE (Esumi et al., 2009). Second, the level of expression was significantly
higher (16-fold) from the conventional transgene (Fig. V-2 and Fig. V-S3).
Together, the qualitative and quantitative differences we observed can explain
the greater abundance of GFP+ cells in Crx>Cre Tg mice compared to Crx>Cre
BAC mice.
We believe the patterns of cumulative transgene expression reflect
dichotomous low-level or leaky activity in proliferating RPCs, and high-level
activity in photoreceptor and bipolar precursors, for four reasons. First, Cre
immunoreactivity closely matched endogenous Crx expression, in adult and
embryonic retinas (Fig. V-2). Second, high-level Crx or Cre expression was
observed only in post-mitotic cells (Fig. V-S7 and data not shown) (Muranishi et
al., 2011). Third, GFP reporter expression was distributed in radial stripes in
BAC transgenic mice (Figs. 2-3 and Suppl Figs. 1,7), suggesting a clonal origin.
Fourth, low concordance (<40%) was observed between R26floxGFP and Z/AP
reporters in the GCL and inner INL of Crx>Math5 BAC and Crx>Cre BAC mice,
but high concordance (~100%) was observed in photoreceptor and bipolar
neurons. This dichotomy contrasts starkly with the uniformly high concordance
observed for a Math5>Cre BAC transgene (Brzezinski et al., 2012), and

205

demonstrates the utility of this approach to distinguish populations with


heterogeneous levels of Cre.
Our analysis suggests that some previous results obtained using
conventional Crx>Cre transgenes should be reinterpreted. Notably, Nishida et al.
(Nishida et al., 2003) ablated Otx2 using a Cre transgene driven by a 12 kb Crx
promoter segment. Paradoxically, this resulted in complete loss of Crx mRNA at
E18.5 (cf. Fig. 4C,D) despite the inherent time delay required for Cre protein
expression, excision of Otx2 genomic sequences, and decay of existing pools of
Otx2 and Crx mRNA and protein (Nagy, 2000). Thus, the ablation of Otx2 in this
experiment must have occurred earlier, in retinal progenitors, well before the
onset of Crx trancription. Likewise, Koike et al. (Koike et al., 2005) conditionally
ablated atypical protein kinase C (aPKC) and observed a major disruption in
retinal organization. On this basis, they concluded that photoreceptors were
critical for proper lamination of the retina. Instead, we believe a more
parsimonious explanation is that aPKC function in RPCs is critical for lamination
and epithelial polarity, as has been demonstrated for other neural progenitors
(Cui et al., 2007; Wodarz et al., 2000).
Ectopic expression of Math5 does not stimulate RGC fate
After establishing the patterns of Crx BAC and Crx 2.4 kb transgenic
expression, we were able to test the effects of massive Math5 overexpression on
the fate trajectory of retinal cells. Our findings show that Math5 overexpression
does not significantly bias RGC fate in the wild-type environment for three
reasons. First, the fraction of RGCs within the GCL was not increased in

206

Crx>Math5 Tg mice at any point during development (Fig. V-4). It is likely that
negative feedback from nascent RGCs (Austin et al., 1995; Belliveau and Cepko,
1999; Waid and McLoon, 1998; Wang et al., 2005; Zhang and Yang, 2001)
restricts Math5 from inducing supernumerary RGCs. Second, the temporal
profile of RGC births was not extended by Crx>Math5 Tg expression (Figs. 5 and
8), despite abundant Crx expression in early post-mitotic precursors during the
post-natal period.
Third, retroviral expression of Math5 did not significantly promote RGC
fate or cell cycle exit in cultured embryonic retinal explants. These results differ
from previous studies in chick and frog, in which overexpression of Math5
orthologs favored RGC fate (Kanekar et al., 1997; Liu et al., 2001). In these
studies, effects on cell cycle dynamics and cell fate could not be completely
isolated. By itself, any experimental manipulation that forces cell cycle exit
during early neurogenesis can cause RPCs to adopt early fates, by effectively
stopping progression of the histogenetic clock (Ohnuma et al., 2002). Indeed,
high-level expression of proneural bHLH factors induces cell cycle exit in vitro
(Farah et al., 2000), and Xath5 overexpression reduces clone size in vivo (Moore
et al., 2002). Furthermore, misexpression of other bHLH factors, such as
Neurod1, during early frog development promotes RGC fate, whereas
overexpression of Xath5 during late developmental stages favors non-RGC fates
(Moore et al., 2002). In mice, Math5 is unlikely to be a major determinant of cell
cycle exit, because it is variably expressed, during or after the terminal division,
and lineage-marked cells do not re-enter the cell cycle in Math5 KO mice

207

(Brzezinski et al., 2012). These disparate gain-of-function results may reflect


differences between species in the timing, level or unique post-translational
regulation of the endogenous Math5 ortholog, or the effective dose of bHLH
protein delivered in these experiments.
Our transgenic and retroviral clone analyses separate cell cycle and fate
effects. In the explant experiments, the level of proviral Math5 expression in
transduced RPCs was not sufficient to promote cell cycle exit (Fig. V-6E) and
RGC fate was not favored (Fig. V-6D). In contrast, the dnMAML retrovirus, which
blocks Notch signaling, drove RPCs out of the cell cycle and increased RGC
abundance, consistent with previous results (Austin et al., 1995; Nelson et al.,
2007; Ohnuma et al., 2002). Some RGC fate effects attributed to Ath5 orthologs
may be explained by premature cell cycle exit. When expressed at eight times
the level of endogenous Crx (Fig. V-2D), ectopic Math5 does not significantly
bias progenitors or post-mitotic precursors towards the RGC fate in a wild-type
environment, supporting its role as a competence factor. However, ectopic
Math5 does favor RGC development in a deficient environment (Math5 KO). We
observed partial rescue of RGC and optic nerve formation in adults and embryos
(Figs. 7-9). These findings are generally consistent with the stimulation of
ganglion cell fate by Math5 transfection in neurosphere cultures (Yao et al., 2007)
or by electroporation of human ATOH7 in Math5 KO retinal explants (Prasov et
al., 2012).
In the wild-type environment, we did observe one relatively minor fate
effect of transgenic Math5. The number of early-born photoreceptors was

208

significantly reduced (2-fold at E15.5) in Crx>Math5 Tg mice. However, adult


retinal histology and photoreceptor morphology were grossly unaffected (Fig. V4E). Furthermore, widespread Math5 expression is not sufficient to stimulate
ectopic Brn3b immunoreactivity (Fig. V-S11), a known downstream target (Liu et
al., 2001; Mu et al., 2005). Finally, the co-expression of Crx and Math5-lacZ in
some progenitors (Fig. V-1) and the persistent expression of transgenic Math5 in
mature adult photoreceptors (Fig. V-2) show that downregulation of Math5 is not
an essential step in the specification of these cell types.
Math5 is not the sole determinant of RGC competence
Our birthdating analyses (Figs. 5-8) provide novel insights into the window
of RGC competence and the role of Math5. The timing of Math5 expression
closely mirrors the RGC birthdating curve and Math5 is required for RGC
competence (Brown et al., 1998; Brown et al., 2001; Brzezinski et al., 2012;
Wang et al., 2001). Thus, in principle, the pattern of Math5 expression may be
the sole factor temporally restricting RGC specification. Indeed, in Gdf11 mutant
mice, an overproduction of RGCs during development is correlated with a
spatiotemporal increase in Math5 expression (Kim et al., 2005). Our results,
however, suggest that prolonged expression of Math5 in Crx>Math5 Tg mice
does not extend the profile of RGC births, even when few nascent RGCs are
present, in the Math5 KO rescue. In the rescued mice, the peak of RGC
birthdates was shifted by approximately two days (Fig. V-8), consistent with the
later onset of Crx expression, but this modest heterochronic effect occured within
the normal envelope for RGC genesis. Furthermore, the profile of residual RGC

209

births in Math5 KO mice closely matches wild-type birthdating curves, except for
the extremely low RGC abundance (Fig. V-8). Together, these findings suggest
that the pattern of Math5 expression is not the sole factor restricting RGC
competence. Instead, a complex network of interactions, including Math5, is
likely to determine the spatiotemporal pattern of RGC genesis. For example, the
envelope of RGC genesis also appears to be shaped by other transcription
factors, Notch signaling, and microRNAs (Elliott et al., 2008; Georgi and Reh,
2010; Silva et al., 2003).
A pioneering model for RGC fate specification
In Crx>Math5 Tg; Math5 KO and Math5 KO mice, the loss of early-born
RGCs (Fig. V-8) is correlated with pathfinding defects in remaining RGCs (Fig. V9 and Fig. V-S10). These observations can be explained in two ways. First, the
residual and rescued RGCs may follow an aberrant RGC differentiation pathway.
In adult Math5 KO mice, residual RGCs form dendritic arbors with normal size
and spacing (Lin et al., 2004), but these cells fail to express Brn3b and Brn3a
(Fig. V-7 and data not shown), which are critical for axon pathfinding, dendritic
stratification, and cytodifferentiation (Badea et al., 2009; Gan et al., 1999). It is
thus possible that Crx>Math5 Tg derived RGCs are intrinsically defective and
express an aberrant set of RNA transcripts. Alternatively, the pathfinding defects
in these RGCs may result indirectly, from a deficiency of early-born ganglion
cells, which may limit the extent of rescue overall. Nascent RGCs are known to
elaborate signals, such as sonic hedgehog (Shh), which promote intraretinal
axon pathfinding generally (Erskine and Herrera, 2007; Oster et al., 2004). In

210

zebrafish, the establishment of early RGC axons is necessary and sufficient for
pathfinding and survival of later RGCs (Pittman et al., 2008), and this community
effect may be widespread in the nervous system (Raper and Mason, 2010). In
Math5 KO mice, residual RGC axons are poorly fasciculated, and often branched
(Fig. V-7), and do not extend radially toward the central retina (Fig. V-S10). In
Crx>Math5 Tg; Math5 KO animals the extent of fasciculation is roughly correlated
with the number of surviving RGCs (data not shown). Thus, isotypic interactions
are likely to be critical for proper pathfinding and fasciculation of the transgenerescued RGCs. These pioneering effects are not limited to intraretinal
pathfinding, as knots of tangled fibers were apparent behind the retinas of
rescued Crx>Math5 Tg and BAC animals (Fig. V-9). These defects appear to be
resolved by apoptosis, although a small number of ganglion cells survive, and
are likely to make synaptic connections in the brain (Triplett et al., 2011). Our
results highlight the robust pathfinding mechanisms that operate in the retina,
and the strong homeostatic mechanisms that balance the ratio of diverse cell
types during development.
Acknowledgements
The authors are grateful to Thom Saunders, Maggie van Keuren and the UM
transgenic animal model core for generating conventional and BAC transgenic
animals; to Sue Tarl, Dellaney Rudolph, Christine Brzezinski and Melinda Nagy
for technical support; to Cheryl Craft for Crx antisera; to Sean Morrison and Ivan
Malliard for the MIG and dnMAML retroviral constructs, respectively; to Mitchell
Gillett for assistance with histology; to Anand Swaroop and Edwin Oh for Crx 2.4

211

kb promoter plasmid; to Chris Edwards, and the UM microscopy and image


analysis laboratory for technical advice. The authors thank Nadean Brown, Chris
Chou, David Turner, and Joe Brzezinski for valuable discussions and critical
reading of the manuscript. This research was funded by National Institutes of
Health (NIH) R01 grant EY14259 (TG). LP was supported by NIH T32 grants
EY13934 and GM07863.

212

Figure V-1. Crx and Math5 (gal) are expressed in overlapping subsets of cells
shortly after cell cycle exit. (A) Schema comparing the termporal expression
patterns and cellular abundance of Crx (red) and Math5 (black) in the retina. (B-C)
Sections from E13.5 (B) or E15.5 (C) Math5-lacZ/+ embryos costained for gal
(Math5-lacZ allele), Crx, and EdU following a 4 hour chase in vivo. Many
cells coexpress Math5 and Crx (arrowheads), and both factors are expressed
in some cells shortly after cell cycle exit (EdU+, arrows). Cells initiating
expression of Crx or Math5 reflect similar populations of progenitors. RPE, retinal
pigmented epithelium; NBL, neuroblastic layer; GCL, ganglion cell layer. Scale bar,
50 m.

213

Figure V-2. Characterization of the Crx>Math5-IRES-Cre and Crx>Cre


conventional and BAC transgenic mice. (A) Map of conventional and
BAC transgenes. In conventional transgenes, Math5-IRES-Cre pA and
control Cre pA cassettes are positioned downstream of the 2.4 kb Crx
promoter fragment, as Crx>Math5 Tg and Crx>Cre Tg, respectively. In
BAC transgenes, equivalent cassettes were precisely inserted in BAC
clone RP23-81H17, at the first ATG of the Crx gene, in exon 2, as
Crx>Math5 BAC and Crx>Cre BAC. (B) Expression patterns of
Crx>Math5 BAC, Crx>Cre BAC, Crx>Math5 Tg, and Crx>Cre Tg mice
carrying R26floxGFP reporters. In adult and embryonic retinas for each
transgene (right, far right), Crx and Cre are coextensive, suggesting
faithful recapitulation of the endogenous Crx pattern. In Crx>Math5 BAC
and Crx>Cre BAC retinas, reporter expression (R26floxGFP) is largely
confined to the photoreceptor and bipolar precursors in the neural retina.
In Crx>Math5 Tg and Crx>Cre Tg, the GFP reporter is widespread
throughout the neural retina, and is apparent in the RPE and ciliary body.
(C) RT-PCR of RNA from P21 Crx>Math5 Tg tissues, P21 wild-type (WT)
retinas, and E14.5 wild-type eyes. The products show total, endogenous
(endo), and transgenic (Tg) Math5 mRNA expression, and actin control.
Endogenous Math5 is expressed in the embyonic retina, and at low levels
in the brain, but is not detected in other tissues (including the adult retina).
In contrast, transgenic Math5 is strongly expressed in P21 retinas, and is
restricted to eye tissue. No PCR products are detected in the absence of
reverse transcriptase ( RT). (D) Taqman qRT-PCR comparing the
relative abundance of Math5 and Crx transcripts in P21 Crx>Math5 BAC
(BAC 60), Crx>Math5 Tg (Tg251), and wild-type retinas. Math5 (a) and
Crx (b) expression levels are normalized to Gapdh, with expression
reported relative to Crx>Math5 BAC (left). Math5 expression is also
reported relative to endogenous Crx levels (right), calculated using triplex
competitive PCR data (k) from Crx>Math5 BAC to normalize expression
levels (Fig. V-S3). Math5 expression in Crx>Math5 Tg retinas is 16-fold
higher than Crx>Math5 BAC retinas, and 8-fold higher than endogenous
Crx. Ex, exon; IRES, internal ribosome entry site; FRT, Flipase
recognition target; on, optic nerve; cb, ciliary body; IS, inner segment;
ONL, outer nuclear layer; INL, inner nuclear layer. Scale bar, 50 m.

214

215

Figure V-3. Widespread Crx>Math5 expression has little effect on cell fate
decisions in the retina. (A-F) Sections from adult Crx>Math5-IRES-Cre BAC
(shortened as Crx>Math5 BAC) or matched Crx>Cre BAC control transgenic
mice carrying R26floxGFP reporters were coimmunostained with GFP and
informative markers. GFP and marker costaining (bottom) and GFP alone (top)
are indicated (arrows). Mller glia were identified by Sox9 (A,B), amacrines by
AP2 (C,D) and horizontal neurons by calbindin (E,F). Arrowheads in E and F mark
calbindin+ horizontal cells that are GFP. (H-I) Flatmounts of adult Crx>Math5 BAC
and Crx>Cre BAC retinas with rhodamine dextran labeled RGCs. Some lineagemarked RGCs (arrows) and displaced amacrine cells (arrowheads) are indicated.
While both transgenes mark essentially all photoreceptors and bipolar cells (Fig.
V-S2), a small fraction of other cell types was also labeled in each case. The
fate spectra are similar, with or without Math5. Rare horizontal cells were
labeled in Crx>Math5 BAC Tg mice, but not in Crx>Cre BAC Tg controls, most
likely due to differences in transgene expression level. Scale bar, 50 m.

216

Figure V-4. Widespread Crx>Math5 expression does not alter RGC abundance or
retinal histology. (A-C) Sections from E13.5 (A), P0 (B) and P22 (C) control or
Crx>Math5-IRES2-Cre transgenic (Crx>Math5 Tg) retinas stained for Brn3a (A,B) or
rhodamine dextran (C) to mark RGCs, and counterstained with DAPI to mark nuclei.
(D) RGC fraction among GCL neurons at P0 and P22. There is no significant
difference in the RGC fraction between transgenic (Tg) and contol retinas. (E) Low
(top) and high (bottom) magnification views of basic fuchsin- and methylene bluestained plastic sections. The retinal histology is similar in Crx>Math5 Tg and control
mice. Scale bar: 50 m in A-C; 100 m in E.

217

Figure V-5. Widespread Crx>Math5 expression does not extend the profile of
RGC births, but decreases the numbers of early-born photoreceptors. (A)
Birthdating of Crx>Math5-IRES-Cre transgenic (Crx>Math5 Tg) and littermate
control retinas. Embryos were exposed to BrdU at E15.5 and analyzed at P22.
There were 2-fold fewer E15.5 birthdated cells in the ONL (photoreceptors) of
Crx>Math5 transgenic animals compared to controls, and corresponding
increases in the INL and GCL. (B) Crx>Math5 Tg pups were similarly exposed to
BrdU at P1 and harvested at P22. Few, if any, GCL cells were labeled with BrdU
in transgenic or control mice. (C) Crx>Math5-IRES-Cre BAC (BAC) embryos were
exposed to EdU at E17.5, and their RGCs were labeled with rhodamine dextran at
P22. No EdU+ RGCs were detected in central flatmounts of BAC or control
retinas. Prolonged transgenic expression of Math5 does not extend the RGC
birthdating profile. Scale bar: 100 m in A-B; 50 m in C.

218

Figure V-6. Retroviral Math5 overexpression does not stimulate RGC fate
or cell cycle exit in retinal explant cultures. (A) Experimental design.
Retinas were explanted from E13.5 embryos, flattened on polycarbonate
membranes, infected at low density with the indicated MSCV retrovirus,
and cultured for 7 days in vitro (DIV). Isolated GFP+ clones were scored
for RGC number by Brn3a immunoreactivity and clone size. (B-C)
Example clones from explants infected with IRES-GFP (B) or Math5IRES-GFP (C) retroviruses. (D) Plot showing the fraction of GFP+ cells
that developed as Brn3a+ RGCs in transduced explants. There was no
significant difference in the RGC fraction of explants transduced with
Math5-IRES-GFP or IRES-GFP control. A modest increase in RGCs was
observed when Notch signaling was autonomously blocked in clones with
the IRES-dnMAML virus. Error bars show binomial standard deviation.
(E) Clone size distribution. There was no difference between Math5IRES-GFP and IRES-GFP explants, but clone size was significantly
reduced in explants infected with IRES-dnMAM. Scale bar, 50 m.

219

220

Figure V-7. Crx>Math5 expression partially rescues RGC fate


specification and optic nerve development in Math5 knockout (KO) mice.
(A-C) Flatmounts of adult retinas from the indicated genotypes stained for
TuJ1 (A, B) to mark axons, and Brn3a (C) to mark cell bodies of mature
RGCs. In Math5 heterozygous mice (het), RGC axons fasciculate and
project to the optic disc (OD), and their cell bodies are reactive for Brn3a.
In Math5 KO mice, RGC axons are very sparse, meander and do not
fasciculate, and their cell bodies not express Brn3a; and the presumptive
optic disc (POD) is not fully developed. The Crx>Math5 Tg and to a
much lesser extent Crx>Math5 BAC transgene is able to partially rescue
RGC density (Brn3a+), pathfinding and fasciculation defects in Math5 KO
mice. (D) Sections from adult retinas of the indicated genotypes, stained
for GFP and Brn3a. In heterozygous mice, the Crx>Math5 Tg marks
~40% of the Brn3a+ RGC population. In contrast, rescued Brn3a+ RGCs
derive exclusively from the Crx>Math5 Tg lineage (arrows). (E) Optic
nerves of adult Math5 het and Crx>Math5 Tg; Math5 KO mice, stained
with H+E. The rescued nerve is much thinner than the control. (F)
Birthdating transgenic and control Math5 KO retinas. Pups were exposed
to a pulse of BrdU at P1 and harvested at P22. Few, if any, GCL cells
were labeled with BrdU in either genotype, indicating that rescued RGCs
are born prior to P1. P, proximal; D, distal. Scale bar, 50 m.

221

222

Figure V-8. RGC birthdates in transgenic, Math5 KO and rescued


animals. Embryos with the four indicated genotypes were exposed to a
single pulse of EdU at E11 and E12, E13.5, E15.5 or E17.5, and are
compared as littermates. (A-D) GCL confocal views of P1 birthdated
retinal flatmounts, stained for Brn3a (magenta) and EdU (green). (E) RGC
birthdating curves for each genotype. No Brn3a+ RGC births were
detected in any genotype after E17. The overall number of RGC births is
substantially reduced in Math5 KO and Crx>Math5 Tg (Tg) rescued mice.
(F) Normalized RGC birthdating curves for each genotype. Because
virtually all RGCs are born between E11 and E17, the total number of
RGCs was summed across the 4 time points for each genotype, and the
RGC birth fraction at each time is plotted relative to this total. The
normalized curves are quite similar. However, very few, if any, Brn3a+
RGCs were born during early neurogenesis (E11-E12) in Math5 knockout
(KO) or rescued (Tg; Math5 KO) mice. In the rescued mice, a larger
fraction of RGCs were born during mid-gestation (E15.5). (G) Brn3a+
RGC density jitter plots for each genotype. Each data point represents a
single eye. The Crx>Math5 transgene (Tg) partially rescues the RGC
deficiency in Math5 KO mice, which have significantly more Brn3a+ RGCs
than Math5 KO controls, although this effect is variable. The Crx>Math5
Tg does not significantly affect RGC density in heterozygotes. Scale bar,
50 m.

223

224

Figure V-9. Survival and generation of late-born RGCs are inhibited in


rescued animals. (A-B) III-tubulin (TuJ1) staining of rescued (Crx>Math5
Tg; Math5 KO) rescued retinas at E15.5 and E17, compared to Math5
heterozygous littermates (het, A,C) or Math5 wild-type (WT) control
(E17.5, B). Math5 KO (left) retinas form a nerve fiber layer, but there is no
clear coalescence of axons into an optic nerve. Crx>Math5 Tg rescued
retinas have thin optic nerves, and these occasionally form axon knots in
the optic stalk (arrowhead, inset). (C-D) Rescued and control sections
stained for cleaved Caspase-3 (arrows) at E16.5 to mark apoptotic cells.
Dying cells primarily reside in the forming ganglion cell layer. Crx>Math5
Tg rescued animals (Tg; KO) exhibit increased levels of apoptosis as
compared to Math5 KO mice, which have much higher levels of cell death
than heterozygous animals. (E) TuJ1 staining of E15.5 Crx>Math5 BAC;
Math5 KO mice, shows partial rescue of optic nerve development and the
appearance of a similar RGC axon knot (arrowhead). Scale bar, 50 m.

225

226

Figure V-S1. Consistent retinal expression patterns for multiple, independent


Crx>Cre Tg and Crx>Cre BAC transgene insertions. Adult retinas from founders or
offspring carrying the R26floxGFP reporter were immunostained for GFP and Cre.
Although each transgenic line gave a similar pattern, some expression was
variegated due to mosaicism (Crx>Cre Tg founder 852 and 855) or epigenetic
silencing (Crx>Cre BAC line 91). (A) In conventional Crx>Cre Tg mice, Cre is
confined to photoreceptors and bipolar cells, while GFP is evident in every retinal
layer. (B) In Crx>Cre BAC mice, Cre and GFP are restricted to the ONL and outer
INL, with occasional stripes of GFP+ cells (line 111). ONL, outer nuclear layer; INL,
inner nuclear layer; GCL, ganglion cell layer. Scale bar, 50 m.

227

Figure V-S2. Crx transgenic and BAC expression patterns in the pineal
gland, retinal pigmented epithelium, and ciliary body. (A) Pineal gland
sections from Crx>Math5 Tg and Crx>Math5 BAC mice stained for
R26floxGFP reporter expression. GFP is evident in the pineal gland, but
absent in surrounding connective tissue. (B) Sections of adult retinal
pigmented epithelium (RPE) from transgenic mice stained for the GFP
reporter. Although the signal is partially obscured by melanin, GFP is
abundant in conventional, but not in BAC transgenic mice. (C)
Immunofluorescence (top) and differential interference contrast (bottom)
images of P1 peripheral retina from a Crx>Math5 Tg; R26floxGFP pup.
GFP expression was apparent in both pigmented and unpigmented ciliary
body (CB) epithelial cells. NR, neural retina.
Scale bar, 50 m.

228

Figure V-S3. Direct comparison of Crx>Math5 BAC and endogenous Crx transcript
levels. (A) Diagram of the triplex competitive RT-PCR strategy. Primers and expected
product sizes are shown within the Crx locus and Crx>Math5 BAC. A common forward
primer labeled with 6-FAM (Ex1) was used in the PCR. Reverse primers in Crx exon 2
(Ex2) or in Math5 (M5) were used in an equal molar ratio. (B) Capillary electrophoresis
profiles of triplex competitive PCR products amplified from Crx>Math5 BAC (top) or
wild-type (WT, bottom) adult retinal RNA templates. In Crx>Math5 BAC retinas (n = 5),
the molar ratio of transgenic Math5 and endogenous Crx transcripts, inferred from the
ratio of PCR products, was 533%, consistent with single-copy expression. This value
(k) was used to normalize qRT-PCR measurements (Fig. V-2D).

229

Figure V-S4. Math5 does not grossly alter secondary fate choices of
photoreceptors and bipolar cells. (A-H) Sections from adult Crx>Math5-IRES-Cre
BAC (Crx>Math5 BAC) or matched Crx>Cre BAC control transgenic (Crx>Cre
BAC) mice carrying the R26floxGFP reporter were coimmunostained for GFP and
indicated photoreceptor or bipolar marker. In the ONL, cone subtypes were
identified by M- or S-opsin staining, and rods by the absence of staining (A-D). In
the INL, bipolar cells were identified by Chx10 staining (E-F), and rod bipolar
subtypes were identified by PKC staining (G-H). Virtually all photoreceptors and
bipolar cells are contained within the Crx lineage (GFP+). There is no apparent
difference in the distribution of these subtypes between transgenes. Scale bar, 50 m.

230

Figure V-S5. Cell fate spectrum of the conventional Crx>Math5 transgene.


Adult Crx>Math5 Tg; R26floxGFP eyes were immunostained for GFP and cell
type specific markers. Descendants of Crx>Math5 Tg expressing cells contribute
to all major cell types, including Mller glia (A, Sox9), amacrines (B, syntaxin),
horizontal neurons (C, calbindin), RGCs (D, retrograde rhodamine dextran
labeling), bipolar cells (E, Chx10), rods (F, rhodopsin) and cones (G, mCar).
Some examples of double-positive cells are marked by arrows. Scale bars, 50
m.

231

232

Figure V-S6. Dual-reporter concordance experiment. (A-B) Lineage analysis


was performed in Crx>Math5-IRES-Cre BAC (A) or Crx>Cre BAC (B) mice
carrying both Z/AP and R26floxGFP reporters. In photoreceptors and bipolar
cells, the concordance between reporters is nearly 100%, suggesting that these
cell types express high levels of Cre. In other cell types, the concordance is low
(<40%), as few cells express GFP and PLAP (arrows). Instead, most lineagemarked cells in the GCL or inner INL express a single reporter (arrowheads).
The labeling of these cell types in the adult retina (Fig. V-3 and Fig. V-S5) can thus
be attributed to low-level or leaky Cre expression. oONL, outer half of the inner nuclear
layer (bipolar domain); iINL, inner half of the inner nuclear layer (amacrine domain).
Scale bar, 50 m.

233

Figure V-S7. The Crx>Math5 BAC transgene is expressed at low levels in


proliferative retinal progenitors. E15.5 Crx>Math5 BAC; R26floxGFP embryos
were pulsed with EdU for 1 hr and costained for the indicated markers. (A)
Projection images of 10 m optical sections show that all Cre+ cells were EdU
indicating that Cre is not expressed at high levels during S-phase. (B) In
contrast, many GFP+ stripes contain EdU+ cells (arrows, inset), consistent
with stochastic Cre expression in proliferating RPCs. (C) A small number of
Brn3b+ RGCs were identified among clustered GFP+ cells (arrow, inset). Scale
bar, 50 m.

234

Figure V-S8. Reduced photoreceptor births at E12.5 and E13.5 in


Crx>Math5 Tg mice. Retinal sections from adult Crx>Math5 Tg and control
mice exposed to a single BrdU pulse at E12.5 (A) or E13.5 (B) were stained
with anti-BrdU and DAPI. Few strongly BrdU+ ONL cells are evident in Crx>Math5
Tg mice, but these are apparent in control mice. Scale bar, 50 m.

235

Figure V-S9. Crx>Math5 Tg mice and control mice exhibit similar levels of
apoptosis throughout development. (A-D) Sections from E13.5, E15.5,
E17.5 or P0 Crx>Math5 transgenic (Tg) mice or control littermates were
stained with DAPI and an antibody to cleaved Caspase-3 to identify
apoptotic cells (arrows). Scale bar, 200 m.

236

Figure V-S10. RGC axons in Math5 KO and transgene-rescued mice exhibit


severe pathfinding defects. (A-C) Confocal Z-stack projections through the
ganglion cell and nerve fiber layers of retinal flatmounts from adult Math5 KO (A),
Crx>Math5 BAC; Math5 KO (B), and Crx>Math5 Tg; Math5 KO (C) mice. Axons
marked by TuJ1 antigen exhibit severe pathfinding defects (arrows), including
branched axon bundles and individual axons, tangled fibers, and nonradial
trajectories. Scale bar, 50 m.

237

Figure V-S11. Brn3b expression in Crx>Math5 mice. (A) Tg and contol


E13.5 embryonic retinas. The abundance of Brn3b+ cells is grossly
similar in the two genotypes. (B) Crx>Math5 BAC and control adult
retinas. There is no Brn3b immunoreactivity in the transgenic ONL and
INL, where Crx>Math5 expression is abundant. Scale bar, 50 m.

238

239
CAGCCCAGGCTTAAAGTCG (Crx ex1)

CGCCGCATGCAGGGGCTGAACACG
ATACCGGAGATCATGCAAGCTGGT
CGCCGCATGCAGGGGCTGAACACG
GCTCTTTTCCAGCCTTCCTT

GTTTGCAGGCCGACTTCATGG (Math5 Tg)

GCCAAGGCATTGACTGAATAG (Crx ex2)

CCTCCTGAATGACGCTAGGA
CCAAACTGGAACAACACTCAACCC
GATTGAGTTTTCTCCCCTAAGACCC
GTACTTGCGCTCAGGAGGAG

GAGTCTGGGACATGTTCAGTT

CCAAACTGGAACAACACTCAACCC

57

60
58
60
58

56

55

Annealing
temp$

1X Masteramp

N/A

1X Masteramp

1X Masteramp

Notes

45-60 sec

These reactions were done using a three-primer PCR, with a common 6-FAM labeled forward primer and equal ratios of reverse primers.

All PCRs had an initial denaturation step (94C x 3 min); followed by 40 cycles of 30-45 sec at 94C for denaturation,
at the noted temperature for annealing, and 1 min at 72C extension; with a final extension step (7 min at 72C)

Single-stranded oligos were used to delete Math5 and IRES sequence by recombineering, to generate the Crx>Cre Tg, and the Crx>Cre
BAC targetting construct

triplex competitive RT-PCR#

endogenous Math5
transgenic Math5
total Math5
actin

CTCTGTTCCTGCTTATTGGGG

Crx BAC genotyping

RT-PCR
RT-PCR
RT-PCR
RT-PCR

ATACCGGAGATCATGCAAGCTGGT

Crx Tg genotyping

CCACAGTCTCTGAAGATCCTGTGATCTCGAAATTCACCATGGGCCCAAAGAAGAAGAGAAAGGTTTCGAA

ss oligo deletion for BACs*

Reverse (antisense primer)

TTCGAAACCTTTCTCTTCTTCTTTGGGCCCATGGTGAATTCGAAATAGGTCCCCTCACACGGGGCGACCT

Forward (sense primer 5'3')

ss oligo deletion for Tg*

Experiment

Table V-S1. Oligonucleotide primers and PCR conditions used in this study

CHAPTER VI
ATOH7 MUTATIONS CAUSE AUTOSOMAL RECESSIVE PERSISTENT
HYPERPLASIA OF THE PRIMARY VITREOUS

Abstract

The vertebrate basic helix-loop-helix (bHLH) transcription factor ATOH7


(Math5) is specifically expressed in the embryonic neural retina and required for
genesis of retinal ganglion cells (RGC) and optic nerves. In Atoh7 mutant mice,
the absence of trophic factors secreted by RGCs prevents development of the
intrinsic retinal vasculature and regression of fetal blood vessels, causing
persistent hyperplasia of the primary vitreous (PHPV). We therefore screened
patients with bilateral optic nerve aplasia (ONA) or hypoplasia (ONH), or
hereditary PHPV, for mutations in ATOH7. We identified a homozygous ATOH7
mutation (p.N46>H) in a large family with an autosomal recessive PHPV disease
trait linked to 10q21, and a heterozygous variant (p.R65>G) in one of five
sporadic ONA patients. High-density SNP analysis also revealed a CNTN4
duplication and an OTX2 deletion in the ONA cohort. Functional analysis of
ATOH7 bHLH domain substitutions, by electrophoretic mobility shift and
luciferase cotransfection assays, revealed that the N46H variant cannot bind
DNA or activate transcription, consistent with structural modeling. The N46H
variant also failed to rescue RGC development in mouse Atoh7 -/- retinal

240

explants. The R65G variant retains all of these activities, similar to wild-type
human ATOH7. Our results strongly suggest that arPHPV is caused by N46H
and is etiologically related to congenital retinal nonattachment (NCRNA). The
R65G allele, however, cannot explain the ONA phenotype. Our study firmly
establishes ATOH7 as a retinal disease gene and provides a functional basis to
analyze new coding mutations.

Introduction
The vertebrate neural retina is a highly ordered laminar structure, which
contains rod and cone photoreceptors, interneurons, specialized glia, and
projection neurons (Rodieck, 1998). During development, these diverse cell
types are generated in a fixed, but overlapping, temporal order from a pool of
multipotent neuroepithelial progenitors (Livesey and Cepko, 2001). At the onset
of retinal neurogenesis, on embryonic day E11 in mice and during the 5th
gestational week in humans, the first neurons exiting the cell cycle differentiate
as retinal ganglion cells (RGCs). Their axons form the optic nerves, which relay
all visual information to the brain.
The adult retina is nourished by two major vascular systems. The central
retinal artery enters the eye along with the optic nerve and supplies the inner
retina, while a tunic of choroidal vessels supplies the outer, photoreceptor layers
(Fruttiger, 2002; Gariano and Gardner, 2005). The developing retina and lens
primordia are perfused by a three-dimensional arterial plexus that branches from
the central hyaloid artery as it emerges from the optic stalk. The hyaloid arcades

241

fill the vitreous and drain into annular vessels at the edge of the optic cup.
During midgestation in humans and early neonatal life in mice, these hyaloid
blood vessels and the associated pupillary membrane regress, and the central
artery remodels to form a two-dimensional branching network that originates at
the optic nerve head (Fruttiger, 2002; Gariano and Gardner, 2005). This
vascular lattice spreads radially across the retinal surface along a scaffold of
migrating astrocytes, which are activated by signals from nascent ganglion cells
(Fruttiger et al., 1996). A second, concentric wave of vessels then branches from
the surface plexus, penetrating to form two deep layers within the mature neural
retina.
The development of retinal neurons is likewise controlled in tandem by
intrinsic transcriptional programs and the microenvironment (Livesey and Cepko,
2001; Ohsawa and Kageyama, 2008; Yang, 2004). As an archetypal instrinsic
factor, the basic helix-loop-helix (bHLH) nuclear protein Atoh7 (Math5) is
transiently expressed in early retinal progenitors (Brown et al., 1998; Brzezinski
et al., 2012) and is necessary for RGC fate specification. Atoh7 -/- mice exhibit a
profound deficiency of RGCs and lack optic nerves (Brown et al., 2001; Wang et
al., 2001). In the absence of RGCs, the mature retinal vasculature fails to form.
Consequently, the hyaloid (fetal) vessels persist, proliferate, and invade the inner
retina (Brzezinski et al., 2003; Edwards et al., 2011). The structure, function, and
spatiotemporal expression of Atoh7 have been highly conserved during
metazoan evolution. Atoh7 was initially identified by its homology to the
invertebrate atonal (ato) and lin32 genes (Brown et al., 1998; Brzezinski et al.,

242

2012), which are required to specify sensory neurons in Drosophila and C.


elegans, respectively (Jarman et al., 1994; Zhao and Emmons, 1995). In
zebrafish, a recessive missense mutation (lakritz) in the Atoh7 ortholog ath5
causes agenesis of RGCs (Kay et al., 2001).
Congenital diseases of the optic nerve and retinal vasculature are
important causes of blindness worldwide, and have some overlapping
pathogenetic features. Optic nerve hypoplasia (ONH) is a common basis for
visual impairment in children, in which the abundance of RGCs is significantly
reduced. Few genetic causes have been identified (McCabe et al., 2011). ONH
is recognized fundoscopically by small-sized optic discs, and may be associated
with neurological or hypothalamic-pituitary dysfunction (Borchert and GarciaFilion, 2008; McCabe et al., 2011). Optic nerve aplasia (ONA) is a much more
severe and rare disorder, in which the optic nerves are essentially absent. Most
ONA patients also have retinal vascular dysgenesis or posterior proliferation, and
some have other ocular or neuroanatomical defects (Blanco et al., 1992; Brodsky
et al., 2004; Lee et al., 1996; Scott et al., 1997). Retinovascular diseases such
as familial exudative vitreoretinopathy (FEVR), retinopathy of prematurity (ROP),
and persistent hyperplasia of the primary vitreous (PHPV), are likewise important
causes of childhood blindness (Gariano and Gardner, 2005). In PHPV, the fetal
hyaloid vessels fail to regress and continue to proliferate (Goldberg, 1997;
Haddad et al., 1978; Reese, 1955; Shastry, 2009). This malformation is typically
sporadic and unilateral, and increases the risk of retinal detachment (Pruett,
1975). PHPV can coexist with FEVR in some eyes. In FEVR, focal avascular or

243

dysplastic regions in the peripheral retina result from incomplete angiogenesis


(Robitaille et al., 2009). Apart from FEVR mutations identified in Wnt pathway
genes (Berger et al., 1992b; Warden et al., 2007), the majority of congenital
retinovascular disease remains unexplained (Gariano and Gardner, 2005).
Given its unique expression pattern and direct role in RGC neurogenesis,
and the importance of RGCs for blood vessel development, ATOH7 mutations
may contribute to the clinical spectrum of congenital optic nerve and
retinovascular disease. Indeed, single-nucleotide polymorphisms (SNPs) in the
ATOH7 locus have been associated with quantitative variation in optic disc area
among healthy individuals with no visual impairment (Khor et al., 2011;
Macgregor et al., 2010; Ramdas et al., 2010), and a similar association has been
reported for glaucoma disease susceptibility (Chen et al., 2012; Fan et al., 2011;
Ramdas et al., 2011). We recently discovered a deletion spanning an upstream
ATOH7 transcriptional enhancer, which causes nonsyndromic congenital retinal
nonattachment (NCRNA) in an Iranian Kurdish population (Ghiasvand et al.,
2011). NCRNA patients lack optic nerves, similar to Atoh7 -/- mice, and have
clinical features that resemble PHPV. In two further studies, ATOH7 point
mutations were described in patients with ONH or vitreoretinal dysplasia, and
may contribute to disease pathogenesis (Khan et al., 2011; Macgregor et al.,
2010).
We have examined the role of ATOH7 in optic nerve aplasia and
hypoplasia, and hereditary PHPV by direct DNA sequencing, and we screened
ONA cases for other genetic defects by whole-genome copy-number variation

244

(CNV) analysis. We identified a basic domain mutation (p.N46>H) that


segregates with autosomal recessive PHPV (MIM 611311) in a previously
characterized pedigree (Khaliq et al., 2001) and a heterozygous mutation
(p.R65>G) in a child with optic nerve aplasia. We critically evaluated the
biochemical and biological properties of these variants, and a second
heterozygous variant (p.A47>T) noted in an optic nerve hypoplasia patient
(Macgregor et al., 2010) using DNA binding, transcriptional activation and RGC
rescue assays. We show that the p.N46>H and lakritz mutations cause a
complete loss of function and are thus likely to be pathogenic, whereas the
p.R65>G substitution has no detectable effect on activity. These studies further
establish ATOH7 mutations as a cause of congenital retinovascular and optic
nerve disease, and contribute to our understanding of bHLH factor function.

Patients and Methods

Ethics statement
All human studies were approved by the University of Michigan
Institutional Review Board and conform to the Declaration of Helsinki (Khaliq et
al., 2001). Mouse studies were approved by the University of Michigan
Committee on the Use and Care of Animals.

Human subjects
PHPV pedigree.

The six-generation Pakistani family with persistent

hyperplastic primary vitreous was described in the original linkage study,

245

including clinical phenotypes of five blind individuals (Khaliq et al., 2001). None
of these patients had documented light perception, and all exhibited gross
nystagmus. In the youngest patient (VI-2), retinal folds were noted in one eye,
but no optic discs were seen during an exam under anesthesia, due to the
presence of a dense white-gray fibrous mass in both vitreae, which obscured the
central fundus. Unfortunately, the optic nerve status of affected individuals could
not be assessed, and they were not available for orbital MRI examination or other
clinical follow-up.
ONA cases.

Patient 1 was blind at birth and has no light

perception. Fundus and MRI exams revealed complete absence of optic nerves,
chiasm, and optic tracts. She developed normally until 8 months, but after 12
months, was at the 5th percentile for height and weight. She exhibited severe
developmental delays in spoken language and other milestones, and at 8 years
of age had persistent difficulties initiating social interactions, an inability to
sustain conversations, poor motor planning and processing, and deficits in
sensory integration and auditory processing. Both parents have normal vision
and cognition. Relevant clinical features of four unrelated ONA patients are
listed in Table VI-S1, including cases 2-4, which were previously reported
(Brodsky et al., 2004; Lee et al., 1996; Scott et al., 1997). Each patient has
bilateral aplasia of the optic nerves, chiasm and tracts, with variable
retinovascular fndings. There was no consanguinity or family history of ocular
disease.

246

ONH cases.

Thirty patients with isolated bilateral optic nerve

hypoplasia and significant visual impairment or nystagmus were selected for


screening. In these cases, the ratio of the horizontal disc diameter to the discmacula distance was less than 0.3 (Borchert et al., 1995). There was no clinical
hypothalamic or pituitary involvement, or brain abnormality based on imaging and
endocrine functional studies.
Mutation screening and SNP genotyping
ATOH7 coding and regulatory elements were amplified from genomic
DNA extracted from whole blood or saliva (Oragene collection kit, Genotek,
Ontario, Canada) samples, using previously described PCR primers and
conditions (Ghiasvand et al., 2011). Products were gel-purified using the Wizard
SV gel system (Promega, Madison, WI) and sequenced on an ABI3730 DNA
Analyzer (Applied Biosystems, Carlsbad, CA) in the University of Michigan DNA
core. For genotyping arPHPV family members, products from ATOH7 coding
amplimer A (Ghiasvand et al., 2011) were digested with EaeI and
electrophoresed through 2.5% agarose gels. For genotyping ONA Patient 1 and
her family members, products from coding amplimer B (Ghiasvand et al., 2011)
were generated using a 32P end-labeled forward primer, digested with BstUI,
electrophoresed through 6% polyacrylamide gels, and exposed to X-ray film. All
variants were compared to high-quality sequence reads in the NHLBI Exome
Variant Database (Project).
Blood DNA from three arPHPV relatives (blind individual, known carrier,
and wild-type sibling), and five sporadic ONA patients and available parents,

247

were analyzed using Illumina Omni1-Quad Infinium BeadChips in the University


of Michigan DNA core. This platform scores 1.1 million informative SNPs,
including 5400 from the arPHPV nonrecombinant interval delimited by D10S1221
and GATA121A08. Copy number variants were annotated using CNV-partition
detection software in GenomeStudio (Ilumina, San Diego, CA), confirmed
manually, and compared against the NCBI Database of Genomic Structural
Variation. All coordinates in this report are based on NCBI reference genome
build 36.1 (hg18).
Sequence alignment and structural modeling
Basic helix-loop-helix domains were aligned as previously described
(Brown et al., 2002). The ATOH7 bHLH domain was modeled based on the
known crystal structure of NeuroD1-E47 complexed with DNA (2QL2) (Longo et
al., 2008) using the SWISS-MODEL server (Schwede et al., 2003). Wild-type
and variant structures were viewed using PyMOL (Schrdinger, LLC).

Plasmid vectors
Full-length human ATOH7 and E47 cDNAs were subcloned in pCS2 or
pCS2-MT vectors (Turner and Weintraub, 1994) with the simian cytomegalovirus
IE94 enhancer/promoter driving expression. For electroporation experiments,
the ATOH7 coding sequence was subcloned in the pUS2-MT-IRES-GFP vector
(Zhang et al., 2012) with an N-terminal 6X-myc epitope tag (MT) and the human
ubiquitin C promoter (UbC) driving expression of a bicistronic transcript that
encodes MT-ATOH7 and GFP.

248

Expression plasmids carrying ATOH7 variants were constructed by sitedirected mutagenesis using the overhanging primer method (Liu and Naismith,
2008) with reagents and conditions noted in Table VI-S2. Reactions were
performed in 1X Pfu Ultra reaction buffer (Stratagene, Santa Clara, CA) with 0.4
M primers, 0.2 mM dNTPs, and 2.5 U of Pfu Ultra HF polymerase.
Masteramp (Epicentre, Madison, WI) was added at 20% (v/v) to melt
secondary structure in the GC-rich ATOH7 cDNA template (Prasov et al., 2010).
Products were digested with DpnI and transformed into E. coli DH5. The
resulting clones were verified by restriction analysis and DNA sequencing.
Electrophoretic mobility shift assays (EMSA)
Nuclear extracts were prepared and electrophoretic mobility shift assays
were performed using established methods (Hellman and Fried, 2007; Wadman
et al., 1997). Briefly, HEK293T cells were transfected in 6 cm plates using
FuGene6 reagent (Roche, Indianapolis, IN) with 1 g plasmid DNA, consisting of
an equal-ratio mixture (1:1) of wild-type or variant pCS2-ATOH7 and pCS2-E47
expression plasmids, or pCS2 empty vector. After 48 hrs, nuclear extracts were
prepared following lysis in cold 10 mM HEPES, 1.5 mM MgCl2, 10 mM KCl, 0.5
mM DTT, 0.5% (w/v) NP40, pH 8.0, and centrifugation at 4000 g for 0.5 min.
The nuclear pellets were resuspended in cold 20 mM HEPES, 1.5 mM MgCl2,
420 mM NaCl, 0.2 mM EDTA, 25% (v/v) glycerol, pH 8.0 and agitated vigorously
for 30 min. The protein content of soluble nuclear lysates was estimated by
Bradford assay (Bradford, 1976). Ten micrograms of each extract was mixed
with 3 U poly[dI-dC] (Sigma, St. Louis, MO) in binding buffer (20 mM HEPES, 50

249

mM KCl, 1 mM EDTA, 25% v/v glycerol, 1 mM DTT, pH 7.6). Double-stranded


oligonucleotide probes were end-labeled with 32P--dCTP using Klenow DNA
polymerase, and added last to the binding reaction. For antibody blocking, 1 g
of mouse anti-ATOH7 immunoglobulin (1A5, Abnova, Taiwan, ROC) was added
immediately to the binding reaction; for cold probe competition, excess unlabeled
oligonucleotide (200 pmol) was introduced. After 20 min incubation at room
temperature, reaction samples were electrophoresed under native conditions
through 4% polyacrylamide gels. The dried gels were exposed overnight to
phosphor screens, which were scanned using the Molecular Imager system
(BioRad, Hercules, CA).

Dual luciferase cotransfection assays


HEK293T cells were transfected with 10-200 ng pCS2-ATOH7 expression
plasmid (wild-type or variant), 600 ng firefly luciferase reporter with 7 tandem
E-box binding sites preceding the promoter (Akazawa et al., 1995; Flora et al.,
2007), and 10 ng DmPol2-Renilla control vector in 12-well plates, using FuGene6
reagent. After 48 hrs, the cells were disrupted with Passive Lysis Buffer
(Promega) and luciferase activity was measured using a Victor-3 model 1420
luminescence plate reader (Perkin-Elmer, Waltham, MA) with Dual-Luciferase
Assay reagents (Promega). Photinus firefly luciferase was normalized to Renilla
luminescence, and the activities of lysates expressing ATOH7 are reported
relative to lysates from cells transfected with empty pCS2 vector. At least two
independent experiments were conducted for each ATOH7 variant, with parallel
transfections in triplicate. Results were compared using Students t-test.

250

Western blot analysis


Soluble proteins from whole cell or nuclear lysates were denatured in 2%
SDS sample buffer, electrophoresed through NuPAGE Novex Bis-Tris 4-12%
polyacrylamide gels, transferred to nitrocellulose membranes, and stained with
Ponceau S. Membranes were blocked in a 1:1 mixture of Tris-buffered saline
(TBS) with 0.2% Tween-20 and Odyssey buffer (LICOR Biosciences, Lincoln,
NE) for 2 hrs, probed with primary antibodies overnight at 4C, washed in TBS
with 0.1% Tween-20, incubated with IRDYE 800CW- or 680LT-conjugated
secondary antibodies (LICOR), and scanned using the Odyssey Imaging System
(LICOR). Primary antibodies were rabbit anti-E47 (1:1000, sc763, Santa Cruz
Biotechnology, Santa Cruz, CA); rabbit anti-ATOH7 (1:500, D01P, Abnova)
(Prasov et al., 2010); mouse anti--tubulin (1:1000, DM1A, Abcam, Cambridge,
MA).

Protein stability assays


HEK293T cells were transfected in 6 cm plates with 2 g of pCS2-ATOH7
expression plasmid and 0.2 g of pUS2-GFP transfection control using FuGene6
reagent. After 24 hrs, cells were treated with cycloheximide (100 g/mL) to block
new protein synthesis (Schneider-Poetsch et al., 2010). Cell pellets were
harvested after 0, 1, 2 or 5 hrs exposure and lysed in RIPA buffer. Stability was
assessed by comparing the ATOH7 signal intensity over time, relative to
endogenous -tubulin, in Western blots of cleared lysates.

251

Retinal explants and electroporation


Retinal explants were prepared as described (Brzezinski et al., 2012), and
manipulated using established DNA electroporation and culture methods
(Donovan and Dyer, 2006; Wang et al., 2002b). Briefly, E13.5 Atoh7 -/- eye cups
(Atoh7tm1Gla allele, (Brown et al., 2001)) were dissected from the sclera and
retinal pigmented epithelium, and bathed with plasmid DNA (1.5 g/l) in Hanks
balanced salt solution (HBSS). Five pulses of 20 V and 50 ms duration,
separated by 950 ms recovery periods, were applied across the retina using a
BTX ECM-830 electroporator and gold-plated tweezer electrodes. Eye cups
were allowed to recover for 5.5 hrs in neurobasal media (Invitrogen, Grand
Island, NY) containing 1X B27 and N2 supplements, glutamine (0.4 mM), BDNF
(50 ng/mL, Peprotech, Rocky Hill, NJ), CNTF (10 ng/mL, Peprotech), penicillin
(50 U/mL), streptomycin (50 g/mL), and gentamicin (0.5 g/mL). After lens
removal, the retinas were flattened onto polycarbonate membranes (0.4 m pore
size, GE Healthcare, Piscataway, NJ) and cultured on Transwell inserts for three
days at 37C at the gas-media interface in a humidified atmosphere containing
5% CO2. After two days of culture, half the media was replaced. On the third
day, explants were fixed in 4% paraformaldehyde for 30 min at room
temperature, washed in PBS, and blocked for 4 hrs in 10% normal donkey
serum (NDS), 1% bovine serum albumin (BSA), 90 g/mL donkey anti-mouse
IgG Fab fragment in PBTx (0.1 M NaPO4 pH 7.3, 0.5% Triton X-100). Whole
explants were then incubated with mouse anti-Brn3a (1:50, 14A6, sc-8429, Santa
Cruz), chicken anti-GFP (1:2000, Abcam), and rabbit anti-3-tubulin/TuJ1

252

antigen (1:500, MRB-435P, Covance, Princeton, NJ) primary antibodies in 3%


NDS, 1% BSA, PBTx overnight at 4C. Explants were washed in PBS, incubated
with Dylight-conjugated secondary antibodies (Jackson Immunoresearch, West
Grove, PA), mounted in Prolong Gold Antifade (Invitrogen), and imaged as
confocal Z-stacks using the Zeiss LSM 510 Meta system.
For each explant, the number of Brn3a+ RGCs was counted within the
transfected population (GFP+) in two high-magnification confocal Z-stacks
extending through the full thickness of GFP+ cells. The RGC fraction (Brn3a+
GFP+ / GFP+) for each ATOH7 construct was averaged from 3-6 explants, in two
series of experiments, and plotted using Prism software (Graphpad). Statistical
comparisons were made using Students t-test.
Competitive genomic PCR and 3 RACE analysis
Endpoints of the CNTN4 tandem duplication were defined by PCR, using
primers with an inverted orientation in the reference genome (Table VI- 2, Fig. VIS4). To co-amplify products from the wild-type (wt) and duplicated (dup) alleles,
competitive genotyping PCRs were performed with a common forward primer in
CNTN4 exon 12 and two allele-specific reverse primers, in introns 12 and 1. The
reactions included 10% (v/v) Masteramp as described (Prasov et al., 2010),
and followed conditions outlined in Table VI-S2.
For 3 RACE (rapid amplification of cDNA ends) analysis, total RNA was
prepared from Epstein-Barr virus-transformed lymphoblastoid cell lines using the
phenol-guanadinium method with Trizol reagent (Invitrogen). cDNA was
generated by reverse transcription (RT) as described (Prasov et al., 2010) with

253

Transcriptor high-fidelity polymerase (Roche). 3 RACE experiments (Frohman


et al., 1988) were performed in two steps, using nested primers and cycling
conditions listed in Table VI-S2.

Results
ATOH7 mutation segregates with PHPV disease
A locus for autosomal recessive persistent hyperplastic primary vitreous
(arPHPV) was mapped in a large consanguineous pedigree to a 13 cM region in
10q21 (Khaliq et al., 2001). This segment contains ATOH7 and the critical region
for nonsyndromic congenital retinal nonattachment (NCRNA), a clinically related
recessive disorder (Ghiasvand et al., 2000) (Fig. VI-1A). In a recent study, we
showed that NCRNA is most likely caused by a 6.5 kb deletion, located 21 kb
upstream from the ATOH7 start site, which removes a transcriptional enhancer
with three evolutionarily conserved noncoding elements. The NCRNA patients
are blind from birth, with no light perception, and have bilateral profusions of
retrolental fibrovascular tissue, similar to arPHPV patients (Khaliq et al., 2001).
They also have optic nerve aplasia, documented by magnetic resonance imaging
(MRI), and early bilateral detachments of the retina, with an apparent tractional
basis. In both diseases, the anterior chambers are shallow and there is
progressive corneal opacification, most likely due to chronic endothelial blood
staining (Brodrick, 1972).
Given the similarity between disease phenotypes, overlap of the linkage
intervals including ATOH7, and potential for shared ancestry between ethnic

254

Kurdish and Pakistani (Baloch) populations, we first tested whether arPHPV


family members carried the chromosome 10q21 NCRNA disease haplotype or
ATOH7 regulatory deletion, by high-density SNP and diagnostic PCR analysis
(Ghiasvand et al., 2011). This excluded haploidentity and revealed no evidence
of a deletion or duplication in the arPHPV disease interval. We then focused our
analysis on the ATOH7 coding region in three first-degree relatives. We
identified a c.136A>C mutation that is homozygous in a blind individual,
heterozygous in his obligate carrier mother, and absent in his unaffected brother
(Fig. VI-1B). This mutation predicts a p.Asn46>His (N46H) amino acid change in
the bHLH domain, and creates a novel EaeI restriction site, which can be used
for DNA genotyping (Fig. VI-1C). We then screened blood DNA of all available
family members by PCR and restriction analysis, and found that this mutation
segregates with the disease as expected (Fig. VI-1D). The N46H variant was not
present in the NHLBI Exome Variant Database among 2462 exomes (Project), in
dbSNP, or in 72 (144 chromosomes) controls with normal vision.

ATOH7 mutation screening in ONA cases


We screened DNA from five unrelated optic nerve aplasia (ONA) cases for
coding or regulatory mutations in the ATOH7 gene, and for deletions or
duplications by Ilumina Omni1-quad SNP genotyping. Two of these patients
(cases 1-2) have additional neurocognitive deficits or anatomical defects, and
three (cases 2-4) have been reported in the literature (Table VI-S1). In Patient 1,
we identified a heterozygous c.193A>G mutation in ATOH7, which predicts a

255

p.Arg65>Lys (R65G) amino acid change (Fig. VI-2). This was confirmed by
restriction analysis of PCR products, and was paternally inherited. A similar
R65G variant was previously reported in a heterozygous Australian child among
12 cases of isolated optic nerve hypoplasia (Macgregor et al., 2010).
Given the potential association between ATOH7 and optic nerve
hypoplasia, we also screened 30 patients with severe isolated ONH for ATOH7
coding mutations. No additional variants were found. Taken together, the
frequency of the ATOH7 R65G among optic nerve aplasia or hypoplasia patients
is estimated to be 2 out of 94 chromosomes. This variant was identified in 7 of
6592 chromosomes from individuals with normal vision, including 0/144 in
control chromosomes, 6/4924 in the exome variant database (Project), and
1/1524 reported by Macgregor et al. (Macgregor et al., 2010). Thus, R65G is
statistically overrepresented among ONA and ONH patients (Fishers exact test,
P = 0.007).
Copy number variant (CNV) analysis
In addition to sequencing of ATOH7 coding and regulatory regions, we
performed high-density SNP analysis to evaluate CNVs among the ONA
patients. In Patient 1, this revealed an 828 kb duplication that is predicted to
disrupt the CNTN4 (contactin) gene (Suppl. Fig 1). We mapped this duplication
by junctional PCR and determined that it was a precise tandem duplication
encompassing exons 2-12 (Fig. VI-S2A-C). Furthermore, in a lymphoblastoid
cell line from Patient 1, this allele can generate a truncated CNTN4 mRNA
transcript by premature polyadenylation, with termination after exon 12 (Fig. VI-

256

S2). The CNTN4 protein (BIG-2) contains an extracellular immunoglobulin


domain, is expressed throughout the brain, and mediates neural cell adhesion
and axon outgrowth (Osterfield et al., 2008; Shimoda and Watanabe, 2009;
Yoshihara et al., 1995). Variants in CNTN4 (MIM 607280) and CNTNAP2
(contactin-associated protein) (MIM 604569) genes, including CNVs, are highly
associated with autism spectrum disorder (Alarcon et al., 2008; Glessner et al.,
2009; Pinto et al., 2010; Roohi et al., 2009). The duplication in Patient 1 is a
novel CNTN4 disruption (Fig. VI-S1B) and relevant to her optic nerve and
neurocognitive phenotypes. The duplicated allele may generate a truncated
protein that could potentially interfere with normal CNTN4 function. Although
CNTN4 is expressed by developing RGC axons in the retinotectal system
(Osterfield et al., 2008; Yoshihara et al., 1995), complete disruption of the
orthologous Cntn4 gene in mice has no gross effect on RGC abundance or optic
nerve development (T. Kaneko-Goto and Y. Yoshihara, personal communication,
and data not shown). Thus, it remains unclear what direct role, if any, the
rearrangement plays in the optic nerve pathology of this patient. However, in the
setting of a CNTN4 rearrangment, blindness is likely to have increased her risk of
neurocognitive dysfunction (Ek et al., 2005; Mukaddes et al., 2007).
In Patient 2, CNV analysis revealed a heterozygous 1.2 Mb deletion
spanning the OTX2 gene (MIM 600037) on chromosome 14q23 (Fig. VI-S3). In
addition to ONA, this patient had anatomical and functional pituitary defects
(Brodsky et al., 2004), consistent with the role of the Otx2 homeodomain protein
in development of the ventral brain, eye and pituitary (Matsuo et al., 1995).

257

Heterozygous OTX2 loss-of-function mutations, including three whole-gene


deletions, have been identified in patients with a variety of severe eye
malformations, including optic nerve aplasia or hypoplasia (Bakrania et al., 2008;
Dateki et al., 2010; Ragge et al., 2005; Schilter et al., 2011; Wyatt et al., 2008),
suggesting a haploinsufficiency mechanism (Chatelain et al., 2006). The optic
nerve phenotype may be attributed to a disruption of optic stalk development
(Martinez-Morales et al., 2001).

In vitro functional analysis of ATOH7 variants


To test the causal link between the ATOH7 variants and disease
phenotypes, we characterized their biochemical effects in vitro using a variety of
structural and functional assays. In addition to N46H and R65G, we included the
p.A47>T variant (A47T) reported by Macgregor et al. (Macgregor et al., 2010) in
our analysis. This allele was identified in a sporadic ONH case, but was not
found in normal control chromosomes (0/5248), including 4924 in the NHLBI
exome variant database (Project). We first evaluated conservation of the
affected residues between species and paralogs by aligning bHLH domains and
comparing mutations in humans, zebrafish, Drosophila and C. elegans (Fig. VI3A). The N46 and A47 residues are contained within the basic (DNA binding)
region, which is the most highly conserved part of the bHLH domain. The R65
residue is less conserved, although it remains a basic amino acid in vertebrates
(Arg or Lys).

258

We next modeled the ATOH7 bHLH domain (Fig. VI-3B) using the known
crystal structure of the NeuroD1-E47 heterodimer bound to DNA (Longo et al.,
2008). In this homology model, the Asn46 side group makes direct contact with
a thymine base in the core E-box recognition site (position 5 of CANNTG).
Introduction of a histidine residue at this position is predicted to significantly
impair DNA binding (Fig. VI-3C, bottom). In contrast, the neighboring Ala47 side
chain does not directly contact DNA, but the threonine substitution may alter
DNA binding properties through conformational effects. The positively charged
Arg65 residue, located at the end of Helix 1, is predicted to interact with the
negatively charged Asp61 side group via an electrostatic bridge, which may
serve to stabilize the helix or limit the flexibility of tertiary interactions (Kumar and
Nussinov, 2002) (Fig. VI-3C, top).
Given the proximity of N46 and A47 residues, and direct contact between
N46 and DNA (Fig. VI-2C), we tested the ability of these variants to bind an Ebox DNA recognition site (Fig. VI-4A). Mammalian plasmid expression vectors
for wild-type or variant ATOH7 proteins were transfected into HEK293T cells, and
the resulting nuclear extracts were compared in an electrophoretic mobility shift
assay (Fig. VI-4B). We also tested a humanized version of the zebrafish lakritz
mutation (p.L56>P), which is believed to cause a complete loss-of-function (Kay
et al., 2001). Because neurogenic bHLH transcription factors bind DNA as
heterodimers with ubiquitous class A bHLH proteins (Murre et al., 1989), we
included an expression plasmid for the binding partner E47 in some
transfections, to augment the low endogenous level of E47 in the HEK293T cell

259

line. Heterodimeric complexes between specific class A bHLH proteins and E47
assemble on DNA and do not form in solution (Wendt et al., 1998). Although this
binding of heterodimers is strongly favored, the E47 protein can also bind DNA
as a homodimer, with a characteristic gel shift pattern corresponding to different
phosphorylated isoforms (Sloan et al., 1996).
The radiolabeled synthetic double-stranded oligonucleotide probe
contained an E-box binding site (CAGGTG) that is optimal for ATOH7 and its
orthologs (Del Bene et al., 2007; Powell et al., 2004) (Fig. VI-4A). In the absence
of added E47, each ATOH7 variant, including wild-type, failed to bind DNA (Fig.
VI-4B). In the presence of wild-type ATOH7, faster migrating ATOH7-E47
complexes are the dominant species bound to DNA. The R65G and A47T
variants formed ATOH7-E47 heterodimers on DNA that were indistinguishable
from wild-type (R65G) or slightly reduced (A47T). These alleles therefore retain
DNA-binding and dimerization activity in vitro. In contrast, extracts containing
N46H or L56P variants formed only E47 homodimeric complexes, indicating that
these mutants are unable to bind DNA. Because the E47 homodimeric
complexes were not disrupted, N46H and L56P do not act as dominant-negative
proteins, like the Id class of bHLH factors (Benezra et al., 1990).
Western analysis confirmed that equivalent levels of ATOH7 and E47
were present in nuclear extracts, suggesting that the ATOH7 variant proteins
have similar stability (Fig. VI-4C). To explore this point further, we performed a
cycloheximide pulse-chase experiment (Zhou, 2004) (Fig. VI-S4). Because each

260

protein variant decayed at approximately the same rate, the observed DNAbinding impairment cannot be attributed to decreased protein stability.
We next tested the transcriptional activity of ATOH7 variants in a
luciferase cotransfection assay using HEK293T cells and an optimized synthetic
reporter, which contains a multimerized E-box sequence (CAGGTG) and minimal
-actin promoter (Akazawa et al., 1995; Flora et al., 2007) (Fig 4D). Wild-type
ATOH7 increased luciferase activity 25-35 fold compared to empty pCS2 vector,
in a dose-dependent manner (Fig. VI-4E). The A47T variant had an intermediate
level of activity compared to wild-type, 60% at the high dose (100 ng) and 20% at
the low dose (20 ng). R65G was indistinguishable from wild-type at the high
dose, but had 60% activity at the low dose. In contrast, N46H and L56P variants
had no detectable activity. Taken together, these results suggest that ATOH7
N46H and L56P are null alleles, whereas A47T and R65G retain significant
function.

Biological rescue of RGC development by ATOH7 variants


To assess function in a biologically relevant system, we compared the
potency of ATOH7 variants in Atoh7 -/- retinal explants. Atoh7 mutant mice have
a >95% reduction in RGCs due to defective fate specification (Brown et al., 2001;
Wang et al., 2001), which is apparent as early as E11 (Brzezinski et al., 2012).
To assay biological rescue of this phenotype, we removed E13.5 eye cups from
Atoh7 -/- embryos and introduced bicistronic expression vectors for ATOH7 and
GFP by ex vivo electroporation (Fig. VI-5A). We then explanted and cultured the

261

retinas for 3 days in vitro (DIV) to allow for RGC differentiation. To assess the
degree of rescue, we immunostained whole explants for GFP, the RGC nuclear
marker Brn3a (Nadal-Nicolas et al., 2009; Xiang et al., 1995), and the
pan-neuronal marker 3-tubulin (TuJ1) (Fig. VI-5B). We then counted the
fraction of transfected (GFP+) cells that developed as ganglion cells (Brn3a+)
(Fig. VI-5C). In explants electroporated with the control vector (GFP only), very
few Brn3a+ cells were detected among the GFP+ cohort (7 3%) and very few
GFP+ TuJ1+ axons were apparent, as expected (Fig. VI-5, Fig. VI-S5A). In
contrast, the majority of cells transfected with the wild-type ATOH7 vector
adopted the RGC fate (72 12% Brn3a+), demonstrating a wide dynamic range
for this assay. Moreover, large bundles of GFP+ TuJ1+ processes, likely
representing RGC axon fascicles, were readily observed in these explants at low
magnification (Fig. VI-S5B). Explants transfected with N46H or L56P variants
failed to develop RGCs (P < 10-5 compared to wild-type, Fig. VI-5C) and could not
be distinguished from the negative control (GFP only), by RGC number (Fig. VI5B,C) or axon density (Suppl. Fig 5C,D). In contrast, R65G and A47T variants
did restore RGC development (Fig. VI-5B,C, Suppl. Fig 5E,F). These results
support our conclusion that N46H and L56P variants are functional null alleles,
while R65G and A47T retain complete or partial biological activity.

Discussion
ATOH7 has an established role RGC development (Brown et al., 2001;
Wang et al., 2001) and a secondary role in retinal vascular development

262

(Brzezinski et al., 2003). ATOH7 mutations are predicted to cause human


blindness (Brown et al., 2002), but the clinical spectrum and molecular
mechanisms remain to be defined (Ghiasvand et al., 2011; Khan et al., 2011;
Macgregor et al., 2010). Given the similarity between mouse Atoh7 phenotypes
and human congenital defects involving the optic nerve or retinal vasculature,
and converging genetic data (Khaliq et al., 2001), we systematically screened a
family with recessive PHPV disease, and a cohort of sporadic patients with major
optic nerve malformations (Table VI-S1), for intragenic or regulatory mutations in
ATOH7. Here, we identified novel coding variants, measured their biochemical
and biological effects, and proved that one variant causes arPHPV.
The N46H mutation causes arPHPV
Our data strongly support a causative role for the p.N46>H mutation in the
arPHPV pedigree (Fig. VI-1) for seven reasons. First, the N46H mutation
segregates with PHPV disease in an autosomal recessive pattern, and the
ATOH7 gene is contained inside the nonrecombinant interval. Second, no other
DNA rearrangement was identified within the disease interval by CNV analysis.
Third, the N46H mutaion removes a highly conserved amino acid residue in the
bHLH domain that directly contacts DNA (Fig. VI-3). Fourth, an orthologous
mutation in Drosphila atonal (ato1) eliminates activity and prevents R8
photoreceptor development in flies (Jarman et al., 1993). Fifth, the mutant
ATOH7 protein had no detectable activity in three independent assays. The
N46H polypeptide is stable, but does not bind DNA or activate transcription via its

263

cognate E-box recognition site (Fig. VI-4), and has no biological function in
restoring RGC development (Fig. VI-5).
Sixth, Atoh7 -/- mice have eye phenotypes that are similar to human
PHPV. In mice, Atoh7 is exclusively expressed by cells in the neural retina
(Brzezinski et al., 2012) and the primary pathology in Atoh7 mutant mice is loss
of RGCs (Brown et al., 2001). However, major vascular defects occur as a
secondary consequence of the ganglion cell deficiency, because RGCs are vital
for proper migration and development of retinal astrocytes (Fruttiger et al., 1996),
which form a scaffold for the growth of intrinsic retinal blood vessels (Fruttiger,
2002). In Atoh7 -/- mice, the intrinsic vasculature fails to develop, and hyaloid
(fetal) vessels persist in the vitreous and proliferate to supply the retina
(Brzezinski et al., 2003; Edwards et al., 2011). These abnormal vessels are
prone to hemorrhage, in the subretinal and intravitreal space, and some
extravasated blood communicates with the anterior chamber (hyphema).
Similarly, in the arPHPV family, all affected family members were blind from birth
and had bilateral retrolental masses (Khaliq et al., 2001). Persistent hyaloid
vessels were clearly observed in the youngest patient. In addition, the adult
patients had anterior chamber pathology, including cataracts, corneal opacity and
anterior synechiae, which may be related to chronic intraocular hemorrhage
(Brodrick, 1972).
Seventh, allelic ATOH7 variants have been identified in three families with
similar retinovascular pathology (Ghiasvand et al., 2011; Khan et al., 2011). In
NCRNA disease, linkage, genomic, and transgenic analysis suggest that a 6.5 kb

264

deletion spanning a remote 5 enhancer impairs ATOH7 transcription, causing


bilateral optic nerve agenesis, abnormal vascularization, and early tractional
detachment of the retina (Ghiasvand et al., 2011). Similarly, in a recent study of
familial vitreoretinal dysplasia, with anterior segment involvement, Khan et al.
identified two ATOH7 mutations (Khan et al., 2011). One of these (E49V) is
close to the N46H variant we discovered, within the basic region (Fig. VI-3A).
They disrupt the core DNA-binding motif (NARER) that is highly conserved
among proneural bHLH factors (Chien et al., 1996; Jarman et al., 1993). The
second mutation causes a frameshift in the ATOH7 polypeptide. Although both
are likely to be deleterious, the effects on protein function have not been
characterized empirically. Together, these ATOH7 mutations suggest a common
pathogenic mechanism, with a primary neuronal basis, for hereditary PHPV,
NCRNA, vitreoretinopathy, and related disorders (Cogan, 1971; Lahav et al.,
1973). Clinical assessment of optic nerve status by orbital magnetic resonance
imaging is thus important for patients with severe retinovascular disease.
ATOH7 in optic nerve aplasia and hypoplasia
The role of ATOH7 in optic nerve aplasia and hypoplasia remains less
clear. We identified an R65G mutation in one ONA case (Patient #1). This allele
is unlikely to cause disease for three reasons. First, functional analysis suggests
that the R65G protein has nearly full activity. The protein is stable, binds DNA,
activates transcription, and promotes RGC development in Atoh7 -/- retinal
explants. The ex vivo rescue analysis (Fig. VI-5) is particularly important,
because this biological assay is a comprehensive test of ATOH7 function.

265

Second, no allelic ATOH7 variant was identified in this patient, in coding or


regulatory sequences. Heterozygous carriers of the N46H null allele or NCRNA
deletion, and Atoh7 +/- mice, have normal optic nerves and no obvious defect in
retinal or vascular development (Brown et al., 2001; Brzezinski et al., 2003;
Ghiasvand et al., 2011). Therefore, a single deleterious allele is not sufficient to
cause disease. Third, although overrepresented in ONH and ONA cases, the
R65G variant was also observed in controls with normal vision, including the
heterozygous father of Patient 1.
Our analysis of the A47T variant previously reported in an ONH patient
(Macgregor et al., 2010) suggests that it is a hypomorphic allele. Activity of A47T
was reduced in all three assays (DNA binding, transactivation, RGC rescue). In
addition, this variant was not found in a large number of normal controls. A47T
may have pathogenetic effects, but is unlikely to be the sole cause of disease
given that only one allele was identified. In principle, these two ATOH7 variants
(A47T and R65G) may lack protein function in a way that was not evaluated by
our in vitro tests. For example, the affected residues may be vital for interactions
between ATOH7 and transcriptional coactivators. Indeed, the A47T mutation
alters one of three residues that confer bHLH specificity to Atonal (ATO) and
Neurogenin (NGN) clades in the Atonal-related bHLH protein family (Quan et al.,
2004).
Implications for Ato class bHLH factor function
Our analysis of the ATOH7 variants provides unique insights into bHLH
domain function. Among proneural bHLH proteins, the NARER motif is highly

266

conserved and important for DNA binding (Atchley and Fitch, 1997; Chien et al.,
1996; Jarman et al., 1993). Our results, however, suggest that an Ala47 Thr
substitution can be tolerated. In contrast, Asn46 makes direct contact with a
DNA base, and is critical for binding and function. In Drosophila, the orthologous
residue is one of three that are altered in the ato1 mutation (Jarman et al., 1994).
Our findings suggest that the Asn Ile substitution at this position is the primary
defect in the ato1 allele.
The ATOH7 R65G variant is predicted to destabilize Helix 1 based on
molecular modeling (Fig. VI-3) and is classified as a potentially damaging variant
based on bioinformatic criteria (Macgregor et al., 2010). However, the R65G
variant had a surprisingly small effect on protein function and stability in vitro, and
on biological function ex vivo. Likewise, the L56P substitution in the zebrafish
lakritz variant behaved unexpectedly. Introduction of a proline at this position is
predicted to disrupt Helix 1 and decrease the overall stability of the protein.
Instead, the humanized lakritz polypeptide was stable (Fig. VI-S1), but failed to
complex with the E-box recognition site (Fig. VI-4), due to impaired dimerization
with E47 or DNA binding.
In principle, variants that fail to bind DNA, such as L56P and N46H, could
act as dominant-negative proteins, similar to the Id class of bHLH factors
(Benezra et al., 1990). Id proteins lack DNA-binding activity, but sequester and
inhibit the function of other bHLH factors (Norton, 2000; Ross et al., 2003). In
contrast, our EMSA results, and the unaffected status of lakritz and arPHPV
heterozygotes, suggest that L56P and N46H mutations are simple null alleles.

267

Together, our genetic, molecular, evolutionary and functional analyses of


ATOH7 mutations connect a variety of clinical and basic observations regarding
hereditary blindness, neurovascular development of the eye, and proneural
bHLH proteins. Our results prove that a bHLH mutation in ATOH7 causes
recessive PHPV, highlight the interdependence of neural and vascular
development, and establish a set of functional tests for analysis of subsequent
ATOH7 variants.

Acknowledgements
The authors are grateful to the patients and families for participating in the
study; to Nathan Vale, Dellaney Rudolph and Susan Tarl for technical support;
to Ingrid Scott, Roberto Warman and Bronwyn Bateman for providing blood
samples from optic nerve aplasia patients; to Bob Lyons, Susan Dagenais and
the University of Michigan DNA core for assistance with the Illumina Omni1Quad
SNP analysis; to Chris Edwards and the UM microscopy and image analysis
laboratory staff for technical advice; to David Turner and Huanqing Zhang for
advice on retinal explants electroporation and for providing the pUS2-MT-IRESGFP vector; to Adriano Flora and Huda Zoghbi for the (E-box)x7 luciferase
reporter plasmid; to Tomomi Kaneko-Goto and Yoshihiro Yoshihara for sharing
Cntn4/BIG-2 mutant mouse tissues; to the Autism Genetic Resource Exchange
(AGRE) consortium for comparative genotyping data; to Donna Martin, Anthony
Antonellis, Chris Chou and David Turner for valuable discussions and critical
reading of the manuscript. The research was funded by grants from The

268

Glaucoma Foundation and the National Institutes of Health (EY14259 and


EY19497) to TG. LP was supported by NIH T32 grants EY13934 and GM07863;
DK and JCS were funded by the Ulverscroft Foundation.

WEB RESOURCES
NHLBI Exome Variant Database

http://evs.gs.washington.edu/EVS/

NCBI Database of Genomic


Structural Variation

www.ncbi.nlm.nih.gov/dbvar/

NCBI Human Reference


Genome Build 36.1

http://www.ncbi.nlm.nih.gov/genome/
assembly/2928/

SWISS-MODEL server

http://swissmodel.expasy.org/

269

Figure VI-1. The ATOH7 p.Asn46>His allele segregates with autosomal recessive
persistent hyperplastic primary vitreous (arPHPV) disease. (A) Minimal region of
shared homozygosity for arPHPV on chromosome 10q21, between D10S1221 and
GATA121A08 (Khaliq et al., 2001). The ATOH7 gene (red) and NCRNA critical
region (Ghiasvand et al., 2000) (yellow) are indicated. The same distal flanking
marker (GATA121A08 or D10S1418) delimits both intervals. (B) ATOH7 sequence
chromatograms from arPHPV family members. The AAC to CAC mutation causes an
Asn46His amino acid substitution (N46H) in the bHLH domain and loss of an EaeI
restriction site (underlined). This variant is heterozygous in the obligate carrier (IV 5)
and homozygous in the affected individual (V 5). (C) Map of the ATOH7 cDNA,
showing the informative PCR product and EaeI restriction fragments. (D) Agarose
gel of EaeI-digested PCR products amplified from arPHPV family members,
numbered as previously reported (Khaliq et al., 2001). Filled and half-filled symbols
show affected indivduals, and carriers assigned by haplotype (Khaliq et al., 2001),
respectively. Carriers have both mutant (270 bp) and wildtype (198 and 72 bp)
alleles, whereas blind individuals have only the mutant allele. NCRNA,
nonsyndromic congenital retinal nonattachment; bHLH, basic helix-loop-helix domain;
wt, wildtype; mut, mutant.

270

Figure VI-2. The ATOH7 p.Arg65>Gly allele in a child with optic nerve aplasia
and developmental delay. (A) Sequence chromatogram showing the
heterozygous AGG to GGG variant in this case (Patient 1). The variant causes
an Arg65Gly (R65G) amino acid substitution, and creates a BstUI restriction site
(underlined) in the PCR product. (B) Map and acrylamide gel autoradiograph
showing wild-type and mutant BstUI-digested PCR products. The forward PCR
primer was end-labeled with 32P, giving wildtype (wt, 208 bp) and mutant (mut,
48 bp) radioactive fragments, respectively. The R65G allele was paternally
inherited.

271

Figure VI-3. Sequence alignment and structural modeling of ATOH7 mutations.


(A) Multispecies amino acid alignment of the ATOH7 bHLH domain and closely
related transcription factors ATOH1 and NEUROD1. Mutations in human,
zebrafish (Danio rerio), Drosophila melanogaster and Caenorhabditis elegans
orthologs, and human NEUROD1, are highlighted (yellow). Two human alleles,
N46H and E49V (Khan et al., 2011) in the basic region, affect invariant Asn46
and Glu49 residues, which are mutated in Drosophila and C. elegans,
respectively. Ala47 is conserved to Drosophila, while the Arg65 is conserved as
a basic amino acid (Arg or Lys) in vertebrates. The zebrafish lakritz mutation
affects an invariant Leu residue, corresponding to L56 in human ATOH7. The
R111L mutation in human NEUROD1 affects an invariant Arg residue and
abolishes DNA binding, causing diabetes (Malecki et al., 1999). (B-C) PyMOL
views of ATOH7 bHLH domains, modeled using the known X-ray crystal
structure of NeuroD1-E47 heterodimers bound to DNA. (B) Low power view
showing locations of the ATOH7 helix, loop and basic domains, bHLH partner
E47, and the E-box DNA binding site. (C) top. High power view of Helix 1
showing the position of the Arg65 residue. Arg65 is predicted to make a chargecharge contact with Asp61, which may stabilize the -helix. bottom. High power
view of the basic region. Asn46 is predicted to directly contact a base (thymine)
within the core E-box recognition site, while Ala47 faces away from the DNA.
arPHPV, autosomal recessive persistent hyperplastic primary vitreous; ONA,
optic nerve aplasia; ONH, optic nerve hypoplasia; VRD, vitreoretinal dysplasia;
T2DM, type II diabetes mellitis.

272

273

Figure VI-4. The arPHPV mutant ATOH7 polypeptide (N46H) does not bind
DNA or activate transcription, while A47T and R65G variants retain these
functions. (A-C) Electrophoretic mobility shift assay (EMSA). (A) Sequence of
EMSA probe radiolabeled with 32P--dCTP (red). The core bHLH recognition site
(E-box) is highlighted. (B) EMSA autoradiogram comparing DNA-binding activity.
HEK293T cells were cotransfected with plasmids encoding wildtype (WT) or
variant ATOH7 proteins, or empty vector (pCS2), with or without heterodimeric
bHLH partner E47. Nuclear extracts were incubated with dsDNA probe and
electrophoresed through a 6% acrylamide gel. In the absence of ATOH7, native
and phosphorylated E47 isoforms (67 kDa) bind DNA as a homodimer, giving a
pattern with two major complexes (E47-E47). In the presence of wildtype
ATOH7 (17 kDa), faster-migrating E47-ATOH7 complexes predominate. ATOH7
variants N46H (arPHPV) and L56P (corresponding to lakritz) fail to bind DNA,
such that only E47 homodimers are observed in these lanes. In contrast, R65G
and A47T variants form ATOH7-E47 heterodimers. Addition of ATOH7 blocking
antibody (Ab) or cold competitor oligo (cc) decreases the amount of probe bound
by the wildtype ATOH7-E47 complex. (C) Western blots of EMSA nuclear
extracts probed with ATOH7 or E47 antibodies show similar levels of these
proteins. (D-E) Luciferase cotransfection assays. (D) The reporter contains 7
tandem E-box sites and a minimal -actin promoter driving expression of firefly
luciferase. (D) Comparison of ATOH7 transcriptional activity. HEK293T cells
were cotransfected with Renilla control and firefly luciferase reporters and
varying doses of ATOH7 expression vectors. Firefly luminescence counts,
normalized to Renilla, are reported as fold activity relative to pCS2 vector. The
N46H and L56P variants have no detectable transcriptional activity, and were
significantly different from wild-type (t-test P < 0.03 for 100 ng and P < 3 x 10-6 for
20 ng). The A47T and R65G variants are not significantly different from wild-type
at the high plasmid dose (100 ng, 60% and P = 0.09 for A47T, 100% and P = 0.77
for R65G), but have reduced activity at the low dose (20 ng, 20% and P < 10-5 for
A47T, 60% and P < 0.001 for R65G). Error bars show the standard deviation of
three replicates from a representative experiment. prom, promoter.

274

275

Figure VI-5. Human ATOH7 R65G and A47T variants rescue ganglion cell
specification in Atoh7 -/- retinal explants, but N46H and L56P mutants do not.
(A) Experimental design. Eye cups from E13.5 Atoh7 -/- embryos were
electroporated ex vivo with a DNA solution containing bicistronic expression
plasmids pUS2-myc-ATOH7-IRES-GFP (encoding wild-type or variant ATOH7
proteins) or pUS2-myc-IRES-GFP (negative control). After a 6-hour recovery
period, retinas were explanted onto polycarbonate membranes, cultured for 3
days in vitro (DIV), fixed and immunostained as wholemount preparations. (B)
Confocal images of ATOH7-transfected retinal explants stained with GFP and
Brn3a antibodies to mark RGCs. In the absence of Atoh7 function, few RGCs
are formed (GFP only). Wild-type (WT), R65G and A47T variants rescue RGC
development in the transfected cell cohort, while N46H and L56P variants do not.
(C) Quantitative analysis of the RGC fraction among transfected cells (GFP+).
Data from two experiments (open and closed circles) are plotted together, with
each point representing a different explant. The mean and SD are indicated.
Each variant was compared to GFP-only or wild-type ATOH7 controls by
Students t-test, with P-values listed below the graph. N46H and L56P mutants
induce significantly fewer RGCs than wild-type ATOH7, and are indistinguishable
from the GFP-only control. IRES, internal ribosome entry site; myc, 6X myc
epitope tag; RGC, retinal ganglion cell. Scale bar, 50 m.

276

277

Figure VI-S1. ONA Patient 1 carries a duplication that disrupts CNTN4. (A)
LogR ratio and B allele frequency plots show Illumina Omni1-Quad SNP
genotyping for the patient and her unaffected mother. The 0.83 Mb duplication
on chromosome 3p26 is shaded in red. The B-allele frequency splitting and logR
ratio increase are characteristic of a 3-to-2 copy gain, which was not maternally
inherited. (B) CNTN4 (contactin-4) genomic region. The duplication in patient 1
(thick red line) spans exons 2-12 and is not present in the database of normal
genomic structural variants. However, other duplications (red) and deletions
(green) overlap CNTN4 in patients with autism spectrum disorder, as reported in
the indicated publications. Human genome coordinates are based on NCBI build
36 (hg18). Mb, megabase.

278

Figure VI-S2. The chromosome 3p26 duplication in Patient 1 is capable of


producing a truncated CNTN4 mRNA. (A) Diagram of the CNTN4
rearrangement, which tandemly duplicates a 832 kb segment encompassing
exons 2-12. Arrows mark the PCR primers used to define the molecular
breakpoint between 5 (blue) and 3 (green) copies and to test the mRNA terminal
structure by 3 RACE (red). (B) Standard duplex (dup) and 3-primer competitive
(compet) PCRs amplify a 579 bp product (dup) that spans the novel junction
between CNTN4 intron 12 and intron 1. Only the wildtype (wt) allele was
amplified in the mother (465 bp). (C) Sequence chromatogram of PCR products
in (B) containing the breakpoint junction. (D,E) 3 RACE analysis of CNTN4
mRNA in lymphoblastoid cell lines, using nested primers in exon 24 (full length)
or exon 12 (truncated). (D) 3 RACE products corresponding to full-length
CNTN4 isoforms were amplified from the patient and her mother. These differ in
size due to variable cDNA priming from the polyA tail and the low abundance of
CNTN4 transcripts in lymphoblastoid RNA. (E) The truncated CNTN4 isoform,
which prematurely terminates in intron 12, was only detected in the patient. (F,G)
Sequence chromatograms of the wild-type (F) and truncated (G) 3 RACE
products. The polyA signal is underlined. In, intron. Ex, exon.

279

280

Figure VI-S3. Chromosome 14q23 deletion in optic nerve aplasia Patient 2


encompasses the OTX2 gene. (A) LogR ratio and B allele frequency plots show
Illumina Omni1-Quad SNP genotyping data. The 1.2 Mb deletion is shaded in
green. The hemizygous region was identified by the pattern of homozygosity in the
B-allele frequency plot and decreased intensity in logR ratio plot. (B) Expanded
view of the critical region. The deletion spans four genes, including OTX2. The
endpoints are located in repetitive DNA, so the exact coordinates and junctional
sequence could not be determined. Parental DNA was not available for analysis.

281

Figure VI-S4. ATOH7 variants have similar protein stability. (A) LICOR dual
fluorescence Western blots of HEK293T cells transfected with wildtype or
variant ATOH7 expression plasmids. Lysates were harvested after 0-5 hours
treatment with cycloheximide (CHX), which blocks synthesis of new proteins.
The level of ATOH7 polypeptide was normalized to endogenous_tubulin, which
has a long half-life. (B) Quantitative analysis of ATOH7 levels in (A) based on
LICOR antibody fluorescence (semilog plot). The variant (red) and wildtype
(black) ATOH7 proteins have equivalent decay kinetics.

282

Figure VI-S5. Low power views of retinal explant rescue experiments.


(A-F) Retinal explants (Fig. VI-5) coimmunostained for GFP and/or
TuJ1 to mark axons. In explants transfected with wild-type ATOH7 (B),
or R65G (E) or A47T (F) variants, ganglion cell axon bundles (arrows)
are abundant. However, in explants expressing GFP only (A), or
ATOH7 L56P (C) or N46H (D) variants, very few axon bundles are evident.
Scale bar, 200 m.

283

284

Sex

Patient
Number

no

yes

no

bilateral micropthalmia,
iris coloboma (OD),
pigmented epithelium
mottling, choroidal
neovascularization

unilateral
microopthalmia and
microcornea(OS);
chorioretinal atrophy;
bilateral iris colobomas

yes

bilateral
microphthalmia;
retinochoroid
depigmentation, absence
of retinal vasculature,
abnormal vessels at area
of optic disc

bilateral
microphthalmia,
bilateral iris coloboma

no

Septo-optic
dysplasia

none

Other ocular
abnormalities

none

none

none

posterior pituitary
ectopia; absence of
pituitary infundibulum;
hypothyroid, pituitary
insufficiency

failure to thrive

Endocrine or pituitary
defects

Table VI-S1. Clinical Features of optic nerve aplasia cases

none

none

none

1.2 MB hemizygous
deletion at 14q23
(includes OTX2)

ATOH7 p.R65>G (het);


828kb tandem dup
CNTN4 ex2-12

developmental delay;
delayed vocalization;
auditory processing
defects

hypoplastic corpus
callosum

Genetic findings

Neurological or brain
defects

this study

Lee et al. 1996

Scott et al. 1997

Brodsky et al. 2004

this study

Reference

285

nested

3RACE CNTN4 in12 upstream

nested

3RACE CNTN4 ex24 upstream

AGACCACTCCCTGCCAATGT

AGACAGCGTTGTTTGGCATC

TGTGTTCCTTTCTTAGTTTGATATGGT

TTTGCTATAGTTTGTCATTTTTGCTT

GGCCACGCGTCGACTAGTAC

GGCCACGCGTCGACTAGTAC

GGCCACGCGTCGACTAGTACTTTTTTTTTTTTTTTTT

3RACE RT

N/A

ACCAACACTGAACTCTTCACCT

CCAGTGTCACAGGAATGTGG

CCACTGGGGAACCACCCCGCGTAAGCGGTCGAAG

GCGGCGGCGCTCGCGCGCGTGGGCCGCCAGGCGCCTG
ATGCGGCGGCGCTCGCGCGTGTTGGCCGCCAGGCGCCT
GCGGTCGAAGGCAGTGTTGGGCCCCTGCATGCGGCGG

Reverse (antisense primer)

CNTN4 WT

TCCAGGTGGTGGGTAAAAGA

CTTCGACCGCTTACGCGGGGTGGTTCCCCAGTGG

ATOH7 p.R65>G mut

CNTN4 duplication

CCTGGCGGCCCACGCGCGCGAGCGCCGCCGCATGCAG
GCCTGGCGGCCAACACGCGCGAGCGCCGCCGCATGC
CGCATGCAGGGGCCCAACACTGCCTTCGACCGCTTAC

Forward (sense primer 5'3')

ATOH7 p.N46>H mut


ATOH7 p.A47>T mut
ATOH7 p.L56>P mut

Experiment

Table VI-S2. Oligonucleotide primers and PCR conditions used in this study

0.2 v/v Masteramp

0.1 v/v Masteramp

95C for 1 min; 30 cycles of [95C


for 1 min, 55C for 1 min, 65C 14
min]; 68C for 10 min
94C for 3 min; 40 cycles of [94C
for 30 sec, 57C for 1 min, 72C 1
min]; 72C 7 min

same as above, except 33 cycles and


60C annealing

94C for 3 min; 25 cycles of [94C


for 30 sec, 58C for 1 min, 72C 1
min]; 72C 7 min

same as above, except 33 cycles

94C for 3 min; 26 cycles of [94C


for 30 sec, 57C for 1 min, 72C 1
min]; 72C 7 min

50C for 1 hr

0.2 v/v Masteramp

Notes

94C for 5 min; 20 cycles of [94C


for 1 min, 57C for 1 min, 68C 6
min]; 68C 10 min

Cycling conditions

CHAPTER VII
DISCUSSION AND FUTURE DIRECTIONS
The results in this thesis provide novel and important insights into the
structure and function of Math5 (Atoh7), the mechanism of RGC fate of
specification, and the role of ATOH7 in human disease. My work has opened up
many new exciting avenues for future directions. My work highlights the
importance of careful experimentation, proper interpretation of findings, and
consideration of alternative hypotheses. In testing new ideas and hypotheses, I
have often uncovered flaws in previous studies, such as the discovery of splicing
of Math5 mRNA (Kanadia and Cepko, 2010) discussed in Chapter II. In each
case, I identify logical explanations for the discrepancies between our findings
and the interpretations of other research groups. Lasting conclusions in science
depend on reproducible results, which come from multiple lines of converging
evidence. As such, there are many future directions that could expand and
strengthen the conclusions presented in this dissertation.
The structure of Math5 (Atoh7)
In Chapters II and VI, we examined the gene, mRNA, and protein
structure of Math5. One observation in these studies was the high G+C content
and secondary structure in the 3 terminus of the Math5 RNA transcript, which
was the likely explanation for the PCR and Northern blotting artifacts observed by

286

Kanadia and Cepko (2010). While this secondary structure is unlikely to disrupt
transcription or RNA processing, Math5 mRNA may be difficult to translate, and
this may provide an additional level of regulation. Indeed, we observed that fulllength Math5 protein could not be produced in E. coli (data not shown). In
eukaryotes, secondary structure and high G+C content may impair translation
(Baim et al., 1985; Kenneson et al., 2001). Further work is necessary to establish
whether translation of Math5 mRNA is a regulated step in retinal progenitors, and
if this step is hypersensitive to ribosome functional status. In principle, impaired
translation of Math5 mRNA could contribute to the severe optic nerve hypoplasia
observed in the setting of Rpl24 riboprotein mutation in Bst/+ mice (Oliver et al.,
2004). This could be tested directly using a Math5-HA knock-in allele (Fu et al.,
2009) in Bst heterozygous mice, or with transfection assays in Bst/+ or wild-type
fibroblasts in vitro.
The fate plasticity of Math5 (Atoh7) cells
In this thesis, I have explored the plasticity of the Math5-expressing cells
(Chapter III and V) and the timing of onset of Math5 and RGC factor expression
relative to the cell cycle (Chapters III and IV). These studies suggest that the
Math5 population is heterogeneous in many aspects, including fate choice and
onset of expression. Indeed, we observed that the Math5>Cre expressing
population can generate all 7 major cell types on the same day during
development. However, it remains unclear whether single Math5+ cells are
multipotent at a given time during development or are heavily biased (restricted)
in their fate selection. Furthermore, it remains to be determined whether the

287

heterogeneity in onset of Math5 expression relative to cell cycle exit has an


impact on the fates chosen by Math5+ cells.
To address these questions, I propose a simple and elegant experiment:
Mosaic Analysis with Double Markers (MADM) of the Math5>Cre lineage (Fig.
VII-1). This technique relies on Cre-mediated interchromosomal recombination,
which leads to reporter expression and sparse marking of clones (Luo, 2007;
Zong et al., 2005) (Fig. VI-1A). In this system, two reciprocally chimeric reporter
genes are knocked into the ubiquitously expressed ROSA26 locus (Zambrowicz
et al., 1997). The two genes are chimeras of GFP and dsRed2, with two halves
separated by an intron containing a loxP site, such that no functional protein is
made until somatic Cre-mediated recombination occurs between these ROSA26
alleles in trans. If recombination occurs during G1, resulting cells are exclusively
yellow, because they contain functional GFP and dsRed2 reporters (Fig. VII-1A).
However, in rare mitotic recombination events during G2 or M phases, two
outcomes are possible. First, one daughter may express both reporters, while
the other daughter is unmarked. Second, each daughter may express a single
functional reporter, such that one daughter is permanently marked by GFP, while
the other is marked with dsRed2. Because these mitotic recombination events
are rare, isolated clones composed of red and green cells can be readily
identified (Fig. VII-1B). This is particularly evident with transiently expressed
Cre-drivers, such as Ngn3>Cre (Desgraz and Herrera, 2009).
I observed that Math5>Cre transgene is expressed only in terminally
mitotic progenitors or in post-mitotic precursors (Chapter III). Thus, we would

288

expect only two clone sizes from MADM analysis using Math5>Cre. Precursors
in which Cre is expressed in post-mitotic (G0) cells would result in a single-cell
clones that are yellow (G1 recombination pattern). In contrast, progenitors in
which Cre is expressed during late S or G2 of the last cell division would
generate two-cell clones with one red and one green cell (X segregation), or
single-cell clones with one yellow and one unmarked cell (Z segregation). If
larger clone sizes are observed, this would necessarily mean that Math5expressing cells continued to cycle. Clone size could also be assessed in Math5
-/- mice, in order to determine whether lineage cells re-enter the cell cycle.
In principle, the fates of daughters in informative two-cell clones could be
concordant or discordant, and may include RGCs or other cell types. Based on
our clonal analysis of explants, many progenitors that give rise to RGCs generate
pairs of this cell type (Chapter IV). These data are seemingly discordant with
previous clonal analysis of frog and rodent retinas (Holt et al., 1988; Turner et al.,
1990), as no definitive two-RGC clones were detected in these studies. Four
major differences between our study and these prior analyses could account for
this discrepancy, and these could largely be reconciled by the in vivo MADM
analysis. First, analyses were carried out in mature adult retinas, after the period
of post-natal RGC culling (Farah and Easter, 2005). Thus, some RGCs among
paired clones likely died during development. Second, due to technical reasons,
progenitors could only be infected at E13.5 or later in the mouse, a time at which
many RGCs have already exited the cell cycle. Thus, RGCs were
underrepresented in clones compared to their contribution to the retina (Jeon et

289

al., 1998; Turner et al., 1990). Third, due to tangential dispersion (Reese and
Galli-Resta, 2002) or cell migration, some RGC clones could not be definitively
assigned. Fourth, species specific differences may result in different clonal
properties between frogs and rodents (Holt et al., 1988).
MADM analysis with Math5>Cre would circumvent each of the limitations.
As Math5>Cre is active at the onset of neurogenesis, in vivo clonal analysis
would follow the earliest cells as they exit mitosis. Coupled with EdU birthdating,
clones could be stratified based on time of cell cycle exit to precisely determine
the fate spectrum Math5+ cells during development. Because clones are sparse
and daughters in two cell clones are marked with different colors, clones could be
unambiguously assigned even in the presence of cell migration. Furthermore,
RGC fate could be scored in flatmounts at P0, prior to the period of post-natal
culling. In the adult, the localization of reporter expression coupled with the
characteristic morphologies and laminar positions of each of the major cell type
would allow for unambiguous scoring of 2-cell clones. MADM analysis could also
be done in Math5 -/- mice, in order to thoroughly evaluate the secondary fate
choices of cells in the Math5 lineage. Taken together, this single experiment
could confirm or challenge our previous findings, and would provide definitive
and novel insights into the fate choices of Math5-expressing cells, the cell cycle
timing of Math5 expression, and the frequency of symmetric RGCs clones.
The non-autonomous role of Math5-expressing cells
Though RGC development requires the function of Math5 (Brown et al.,
2001; Wang et al., 2001), we found that only 55% of RGCs are contained in the

290

Math5 lineage (Chapter III). Furthermore, an even smaller percentage (only 2030%) of the earliest born cohort of RGCs expresses Math5, yet nearly all of these
cells are lost in Math5 -/- mice. Together, these results suggest a nonautonomous role of Math5 in the specification of RGCs. It is likely that a
secreted or cell surface factor is made by Math5 cells, but not by neighboring
neurogenic cells in the early retina.
Two sets of experiments could be used to confirm this role of Math5 and
also to identify the precise factors involved. First, analysis of chimeric mice
carrying a mix of Math5 -/- and reporter-marked wild-type cells could provide
confirmation that Math5 cells have a non-autonomous role in RGC development.
Specifically, in chimeras composed predominantly of wild-type cells, we could
determine whether Math5 -/- cells are more likely to adopt the RGC fate. In
reciprocal chimeras, we would expect the opposite results, i.e. wild-type cells in a
Math5 -/- environment, cannot form mature RGCs. Alternatively, Math5 could be
conditionally deleted or overexpressed using a mosaic Cre, such as Chx10>Cre
(Rowan and Cepko, 2004). Second, microarray analysis could be conducted to
determine the gene expression profiles of flow-sorted early Math5-expressing
progenitors, using Math5>GFP (Hufnagel et al., 2007) or ATOH7-3034>mCherry
(Ghiasvand et al., 2011) transgenic mice. These profiles could be compared in
Math5 -/- and +/+ backgrounds, to determine what specific factors are lost in
Math5 -/-. This analysis would provide candidate genes and factors that nonautonomously regulate RGC development in the early E11.5-E12.5 retina.

291

The RGC fate decision and the role of Brn3b


In recent years, clear pathways have been established to direct human
embryonic stem cells towards the retinal cell fates (Lamba et al., 2006; Reh et
al., 2010). Furthermore, three-dimensional culture has been used successfully to
make a primordial optic cup (Eiraku et al., 2011). In addition, photoreceptor
precursors derived from transgenic mice or human embryonic stem cells have
been shown to integrate into the retina and restore visual function in animal
models with photoreceptor degeneration (Lamba et al., 2009; MacLaren et al.,
2006).
Generating retinal ganglion cells at high density and purity from embryonic
stem cells has proven to be much more difficult. One missing element is a
master regulatory gene for RGC development. Math5 was a logical candidate for
this role, as it occupies a central node atop the RGC differentiation cascade (Mu
et al., 2005). However, the work presented in this dissertation suggests that
Math5 acts purely as a competence factor (Chapters III and V), and thus does
not serve as a master regulator. The factors that bias competent progenitors
towards the RGC fate remain to be determined. One potential regulator may be
Brn3b (Pou4f2). I have shown that this factor is expressed in RGCs and not
developing amacrine and horizontal cells (Chapter IV), and previous work has
demonstrated that it is critical for terminal differentiation of RGCs and inhibition of
amacrine fate (Erkman et al., 1996; Gan et al., 1996; Qiu et al., 2008; Xiang,
1998). In principle, Brn3b may substantially bias RGC fate in competent mouse
progenitors, similar to that observed in chick retroviral over-expression studies

292

(Liu et al., 2000). However, definitive Brn3b lineage and over-expression studies
in the mouse are lacking. To prove that Brn3b is a marker of committed RGCs,
the descendents of Brn3b cells could be traced using a Brn3b>Cre BAC
transgenic mouse, similar to the tracing experiments presented in Chapter III. It
is expected that lineage-marked cells only contain ganglion cells. To establish
the role of Brn3b in biasing RGC fate, Brn3b could be expressed in the Math5
pattern, using a Math5>Brn3b-IRES-Cre BAC transgene. In this way, lineal
descendents of Brn3b over-expressing cells could be followed by crossing to
floxed reporters. It is expected that Brn3b would bias Math5 lineage cells
towards RGCs and suppress amacrine fates. Together, these studies would
determine if Brn3b is sufficient for stimulating ganglion cell production in
competent progenitors and if it is a master regulator of RGC development.
Testing the pioneering model of RGC fate
In Chapter V, I thoroughly characterized the birthdates of ganglion cells in
Crx>Math5 Tg; Math5 -/- mice. I found that high-level expression of Math5 did
not extend the profile of RGC births, and that early born ganglion cells were
largely lost in these mice. This event was correlated with pathfinding defects,
RGC death, and incomplete rescue. These results support a pioneering model of
RGC fate in the mouse, as is evident in zebrafish (Pittman et al., 2008).
However, causation could not be definitely established due to confounding
variables in the experiment. Thus, a cleaner experiment could directly test the
pioneering model of RGC development in mice, and determine the critical
embryonic times for this pioneering. In principle, the critical window of pioneering

293

RGC axons may be related to the closure of the optic fissure, which occurs
during early development (Barishak, 1992; Gregory-Evans et al., 2004). This
critical time period could be determined using an inducible Math5 expression
system in the context of a Math5 -/- genotype, i.e. with a dual transgenic Tet-On
system (reviewed in (Corbel and Rossi, 2002)). While this may be cumbersome,
Math5 expression, and thus the window of RGC births, could be precisely
controlled and delayed. It is expected that induction at E11.5 would result in
complete rescue, while induction at later developmental time points (e.g. E14.5)
would rescue initial specification, but not terminal differentiation of RGCs.
The role of ATOH7 (Math5) in human disease
Data presented in Chapter VI, together with recent studies (Ghiasvand et
al., 2011; Khan et al., 2011), have established a role for ATOH7 in congenital
neurovascular diseases of the retina. Causative mutations in ATOH7 were
identified in a group of related developmental disorders, including non-syndromic
congenital retinal non-attachment (NCRNA, Ghiasvand et al., 2011), vitreoretinal
dysplasia (VRD, Khan et al., 2011), and persistent hyperplastic primary vitreous
(PHPV, Chapter VI). These diseases all share common pathogenetic features
and an overlapping clinical spectrum of symptoms. The common genetic feature
among patients is a homozygous loss-of-function mutation in ATOH7. Thus, the
underlying cause of disease is likely the agenesis of the optic nerve. This was
not appreciated in many of these patients, because retinovascular disease
prevented visualization of the fundus (Ghiasvand et al., 1998; Khaliq et al.,
2001). Given the wide spectrum of symptoms in among NCRNA, VRD, and

294

PHPV, visualization of optic nerves by magnetic resonance imaging should be


implemented in patients with congenital retinovascular disease. The prevalence
of ATOH7 mutations in PHPV, NCRNA, and VR has not been thoroughly
established. However, ATOH7 is a good genetic candidate for these diseases,
as well as related disorders such as familial exudative vitreoretinopathy (FEVR)
where no mutations have been found in Wnt signaling pathway genes (Warden
et al., 2007). Thus, patients presenting with retinal vascular disease and optic
nerve hypoplasia or hypoplasia, in the absence of other CNS or pituitary defects,
should be screened for ATOH7 mutations.
Concluding remarks
In summary, I have presented seminal findings about the role of mouse
Math5 in retinal development, and the role of its human ortholog ATOH7 in
disease. My thesis has clearly elucidated the structure and expression profile of
the Math5 gene. I have precisely timed the RGC fate commitment step, which
revealed a striking heterogeneity among retinal progenitors. Furthermore, my
work has definitively established that Math5 is not sufficient for RGC fate
specification. I have shown that Math5-expressing cells have a non-autonomous
role in RGC development, and suggested that pioneering ganglion cells are
important for the survival and pathfinding of late born RGCs. Additionally, I have
identified causative mutations in ATOH7 in PHPV, and have shown that
mutations in ATOH7 are not a common cause of isolated optic nerve hypoplasia
or aplasia. Together, these findings extend our understanding of RGC
development and provide novel insights that may be important for diagnostic

295

testing and the development of ganglion cell replacement therapies in diseases


of the optic nerve.

296

Figure VII-1. Outline of the Mosaic Analysis of Double Markers (MADM) strategy.
(A) The MADM technique consists of two reciprocally chimeric reporters.
Recombination can occur in G2, and be resolved by X or Z segregation. In X
segregation, a single cell is double-colored. In Z segregation, both daughter cells are
labeled, with one functional reporter active in each (one green and one red cells).
Recombination can also occur in G1 or post-mitotic cells, which generates doublecolored cells. (B) Example of sparse-labeling in pancreatic islet cells using MADM
strategy with Ngn3>Cre driver. Images adapted from Zong et al., 2005 and Desgraz
and Herrera, 2009.

297

REFERENCES
Abu-Amero, K. K. (2011). Leber's Hereditary Optic Neuropathy: The
Mitochondrial Connection Revisited. Middle East Afr J Ophthalmol 18, 17-23.
Adler, R. and Hatlee, M. (1989). Plasticity and differentiation of embryonic
retinal cells after terminal mitosis. Science 243, 391-3.
Ahmad, I., Dooley, C. M. and Polk, D. L. (1997). Delta-1 is a regulator of
neurogenesis in the vertebrate retina. Dev Biol 185, 92-103.
Akazawa, C., Ishibashi, M., Shimizu, C., Nakanishi, S. and Kageyama, R.
(1995). A mammalian helix-loop-helix factor structurally related to the product of
Drosophila proneural gene atonal is a positive transcriptional regulator expressed
in the developing nervous system. J Biol Chem 270, 8730-8.
Akimoto, M., Cheng, H., Zhu, D., Brzezinski, J. A., Khanna, R., Filippova, E.,
Oh, E. C., Jing, Y., Linares, J. L., Brooks, M. et al. (2006). Targeting of GFP to
newborn rods by Nrl promoter and temporal expression profiling of flow-sorted
photoreceptors. Proc Natl Acad Sci U S A 103, 3890-5.
Alarcon, M., Abrahams, B. S., Stone, J. L., Duvall, J. A., Perederiy, J. V.,
Bomar, J. M., Sebat, J., Wigler, M., Martin, C. L., Ledbetter, D. H. et al.
(2008). Linkage, association, and gene-expression analyses identify CNTNAP2
as an autism-susceptibility gene. Am J Hum Genet 82, 150-9.
Alexiades, M. R. and Cepko, C. (1996). Quantitative analysis of proliferation
and cell cycle length during development of the rat retina. Dev Dyn 205, 293-307.
Allingham, R. R., Liu, Y. and Rhee, D. J. (2009). The genetics of primary openangle glaucoma: a review. Exp Eye Res 88, 837-44.
Alonso, S., Minty, A., Bourlet, Y. and Buckingham, M. (1986). Comparison of
three actin-coding sequences in the mouse; evolutionary relationships between
the actin genes of warm-blooded vertebrates. J Mol Evol 23, 11-22.
Altschuler, D. M., Turner, D. L. and Cepko, C. L. (1991). Specification of cell
type in the vertebrate retina. In Cell Lineage and Cell Fate in Visual System
Development, (ed. D. M. K. Lam and C. J. Shatz), pp. 37-58. Cambridge, MA:
MIT Press.

298

Altshuler, D. M., Turner, D. L. and Cepko, C. L. (1991). Specification of cell


type in the vertebrate retina. In Cell Lineage and Cell Fate in Visual System
Development., (ed. D. M. K. Lam and C. J. Shatz), pp. 37-58. Cambridge, MA:
MIT Press.
Andrews, B. J., Proteau, G. A., Beatty, L. G. and Sadowski, P. D. (1985). The
FLP recombinase of the 2 micron circle DNA of yeast: interaction with its target
sequences. Cell 40, 795-803.
Applebury, M. L., Antoch, M. P., Baxter, L. C., Chun, L. L., Falk, J. D.,
Farhangfar, F., Kage, K., Krzystolik, M. G., Lyass, L. A. and Robbins, J. T.
(2000). The murine cone photoreceptor: a single cone type expresses both S and
M opsins with retinal spatial patterning. Neuron 27, 513-23.
Ashery-Padan, R. and Gruss, P. (2001). Pax6 lights-up the way for eye
development. Curr Opin Cell Biol 13, 706-14.
Atchley, W. R. and Fitch, W. M. (1997). A natural classification of the basic
helix-loop-helix class of transcription factors. Proc Natl Acad Sci USA 94, 5172-6.
Austin, C. P., Feldman, D. E., Ida, J. A., Jr. and Cepko, C. L. (1995).
Vertebrate retinal ganglion cells are selected from competent progenitors by the
action of Notch. Development 121, 3637-50.
Azuma, N., Yamaguchi, Y., Handa, H., Tadokoro, K., Asaka, A., Kawase, E.
and Yamada, M. (2003). Mutations of the PAX6 gene detected in patients with a
variety of optic-nerve malformations. Am J Hum Genet 72, 1565-70.
Badea, T. C., Cahill, H., Ecker, J., Hattar, S. and Nathans, J. (2009). Distinct
roles of transcription factors brn3a and brn3b in controlling the development,
morphology, and function of retinal ganglion cells. Neuron 61, 852-64.
Baim, S. B., Pietras, D. F., Eustice, D. C. and Sherman, F. (1985). A mutation
allowing an mRNA secondary structure diminishes translation of Saccharomyces
cerevisiae iso-1-cytochrome c. Mol Cell Biol 5, 1839-46.
Baker, K. E. and Parker, R. (2004). Nonsense-mediated mRNA decay:
terminating erroneous gene expression. Curr Opin Cell Biol 16, 293-9.
Bakrania, P., Efthymiou, M., Klein, J. C., Salt, A., Bunyan, D. J., Wyatt, A.,
Ponting, C. P., Martin, A., Williams, S., Lindley, V. et al. (2008). Mutations in
BMP4 cause eye, brain, and digit developmental anomalies: overlap between the
BMP4 and hedgehog signaling pathways. Am J Hum Genet 82, 304-19.
Barishak, Y. R. (1992). Embryology of the eye and its adnexae. Dev Ophthalmol
24, 1-142.

299

Barnstable, C. J., Hofstein, R. and Akagawa, K. (1985). A marker of early


amacrine cell development in rat retina. Brain Res 352, 286-90.
Barton, K. M. and Levine, E. M. (2008). Expression patterns and cell cycle
profiles of PCNA, MCM6, cyclin D1, cyclin A2, cyclin B1, and phosphorylated
histone H3 in the developing mouse retina. Dev Dyn 237, 672-82.
Bassett, E. A., Pontoriero, G. F., Feng, W., Marquardt, T., Fini, M. E.,
Williams, T. and West-Mays, J. A. (2007). Conditional deletion of activating
protein 2alpha (AP-2alpha) in the developing retina demonstrates non-cellautonomous roles for AP-2alpha in optic cup development. Mol Cell Biol 27,
7497-510.
Baye, L. M. and Link, B. A. (2008). Nuclear migration during retinal
development. Brain Res 1192, 29-36.
Belliveau, M. J. and Cepko, C. L. (1999). Extrinsic and intrinsic factors control
the genesis of amacrine and cone cells in the rat retina. Development 126, 55566.
Benezra, R., Davis, R. L., Lockshon, D., Turner, D. L. and Weintraub, H.
(1990). The protein Id: a negative regulator of helix-loop-helix DNA binding
proteins. Cell 61, 49-59.
Berger, W., Meindl, A., van de Pol, T. J., Cremers, F. P., Ropers, H. H.,
Doerner, C., Monaco, A., Bergen, A. A., Lebo, R., Warburgh, M. et al.
(1992a). Isolation of a candidate gene for Norrie disease by positional cloning.
Nat Genet 2, 84.
Berger, W., van de Pol, D., Warburg, M., Gal, A., Bleeker-Wagemakers, L.,
de Silva, H., Meindl, A., Meitinger, T., Cremers, F. and Ropers, H. H. (1992b).
Mutations in the candidate gene for Norrie disease. Hum Mol Genet 1, 461-5.
Berget, S. M. (1995). Exon recognition in vertebrate splicing. J Biol Chem 270,
2411-4.
Berget, S. M., Berk, A. J., Harrison, T. and Sharp, P. A. (1978). Spliced
segments at the 5' termini of adenovirus-2 late mRNA: a role for heterogeneous
nuclear RNA in mammalian cells. Cold Spring Harb Symp Quant Biol 42 Pt 1,
523-9.
Berk, A. J. and Sharp, P. A. (1977). Sizing and mapping of early adenovirus
mRNAs by gel electrophoresis of S1 endonuclease-digested hybrids. Cell 12,
721-32.
Bertrand, N., Castro, D. S. and Guillemot, F. (2002). Proneural genes and the
specification of neural cell types. Nat Rev Neurosci 3, 517-30.

300

Biagini, G., Baldelli, E., Longo, D., Contri, M. B., Guerrini, U., Sironi, L.,
Gelosa, P., Zini, I., Ragsdale, D. S. and Avoli, M. (2008). Proepileptic influence
of a focal vascular lesion affecting entorhinal cortex-CA3 connections after status
epilepticus. J Neuropathol Exp Neurol 67, 687-701.
Blackburn, D. C., Conley, K. W., Plachetzki, D. C., Kempler, K., Battelle, B.
A. and Brown, N. L. (2008). Isolation and expression of Pax6 and atonal
homologues in the American horseshoe crab, Limulus polyphemus. Dev Dyn
237, 2209-19.
Blackshaw, S., Harpavat, S., Trimarchi, J., Cai, L., Huang, H., Kuo, W. P.,
Weber, G., Lee, K., Fraioli, R. E., Cho, S. H. et al. (2004). Genomic analysis of
mouse retinal development. PLoS Biol 2, E247.
Blanco, R., Salvador, F., Galan, A. and Gil-Gibernau, J. J. (1992). Aplasia of
the optic nerve: report of three cases. J Pediatr Ophthalmol Strabismus 29, 22831.
Blanks, J. C. and Johnson, L. V. (1983). Selective lectin binding of the
developing mouse retina. J Comp Neurol 221, 31-41.
Boije, H., Edqvist, P. H. and Hallbook, F. (2009). Horizontal cell progenitors
arrest in G2-phase and undergo terminal mitosis on the vitreal side of the chick
retina. Dev Biol 330, 105-13.
Borchert, M. and Garcia-Filion, P. (2008). The syndrome of optic nerve
hypoplasia. Curr Neurol Neurosci Rep 8, 395-403.
Borchert, M., McCulloch, D., Rother, C. and Stout, A. U. (1995). Clinical
assessment, optic disk measurements, and visual-evoked potential in optic nerve
hypoplasia. Am J Ophthalmol 120, 605-12.
Bovolenta, P., Mallamaci, A., Briata, P., Corte, G. and Boncinelli, E. (1997).
Implication of OTX2 in pigment epithelium determination and neural retina
differentiation. J Neurosci 17, 4243-52.
Bradbury, E. M. (1992). Reversible histone modifications and the chromosome
cell cycle. Bioessays 14, 9-16.
Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye binding. Anal
Biochem 72, 248-54.
Brett, D., Pospisil, H., Valcarcel, J., Reich, J. and Bork, P. (2002). Alternative
splicing and genome complexity. Nat Genet 30, 29-30.

301

Brincat, J. L., Pfeiffer, J. K. and Telesnitsky, A. (2002). RNase H activity is


required for high-frequency repeat deletion during Moloney murine leukemia
virus replication. J Virol 76, 88-95.
Bringmann, A., Pannicke, T., Grosche, J., Francke, M., Wiedemann, P.,
Skatchkov, S. N., Osborne, N. N. and Reichenbach, A. (2006). Muller cells in
the healthy and diseased retina. Prog Retin Eye Res 25, 397-424.
Brodrick, J. D. (1972). Corneal blood staining after hyphaema. Br J Ophthalmol
56, 589-93.
Brodsky, M. C., Atreides, S. P., Fowlkes, J. L. and Sundin, O. H. (2004). Optic
nerve aplasia in an infant with congenital hypopituitarism and posterior pituitary
ectopia. Arch Ophthalmol 122, 125-6.
Brodsky, M. C. and Glasier, C. M. (1993). Optic nerve hypoplasia. Clinical
significance of associated central nervous system abnormalities on magnetic
resonance imaging. Arch Ophthalmol 111, 66-74.
Brown, N. L., Dagenais, S. L., Chen, C. M. and Glaser, T. (2002). Molecular
characterization and mapping of ATOH7, a human atonal homolog with a
predicted role in retinal ganglion cell development. Mamm Genome 13, 95-101.
Brown, N. L., Kanekar, S., Vetter, M. L., Tucker, P. K., Gemza, D. L. and
Glaser, T. (1998). Math5 encodes a murine basic helix-loop-helix transcription
factor expressed during early stages of retinal neurogenesis. Development 125,
4821-33.
Brown, N. L., Patel, S., Brzezinski, J. and Glaser, T. (2001). Math5 is required
for retinal ganglion cell and optic nerve formation. Development 128, 2497-508.
Brzezinski, J. (2005). The role of Math5 in retinal development. In Human
Genetics, vol. Ph. D. (ed., pp. 236. Ann Arbor: University of Michigan.
Brzezinski, J. A., Brown, N. L., Tanikawa, A., Bush, R. A., Sieving, P. A.,
Vitaterna, M. H., Takahashi, J. S. and Glaser, T. (2005). Loss of circadian
photoentrainment and abnormal retinal electrophysiology in Math5 mutant mice.
Invest Ophthalmol Vis Sci 46, 2540-51.
Brzezinski, J. A. and Glaser, T. (2004). Math5 establishes retinal ganglion cell
competence in postmitotic progenitor cells. Invest Ophthalmol Vis Sci 45, 3422
E-abstract.
Brzezinski, J. A., Kim, E. J., Johnson, J. E. and Reh, T. A. (2011). Ascl1
expression defines a subpopulation of lineage-restricted progenitors in the
mammalian retina. Development 138, 3519-31.

302

Brzezinski, J. A., Lamba, D. A. and Reh, T. A. (2010). Blimp1 controls


photoreceptor versus bipolar cell fate choice during retinal development.
Development 137, 619-29.
Brzezinski, J. A., Prasov, L. and Glaser, T. (2012). Math5 defines the ganglion
cell competence state in a subpopulation of retinal progenitor cells exiting the cell
cycle. Dev Biol In press.
Brzezinski, J. A., Schulz, S. M., Crawford, S., Wroblewski, E., Brown, N. L.
and Glaser, T. (2003). Math5 null mice have abnormal retinal and persistent
hyaloid vasculatures. Dev Biol 259, 394.
Buck, S. B., Bradford, J., Gee, K. R., Agnew, B. J., Clarke, S. T. and Salic, A.
(2008). Detection of S-phase cell cycle progression using 5-ethynyl-2'deoxyuridine incorporation with click chemistry, an alternative to using 5-bromo2'-deoxyuridine antibodies. Biotechniques 44, 927-9.
Buono, L. M., Foroozan, R., Sergott, R. C. and Savino, P. J. (2002). Is normal
tension glaucoma actually an unrecognized hereditary optic neuropathy? New
evidence from genetic analysis. Curr Opin Ophthalmol 13, 362-70.
Carelli, V., Ross-Cisneros, F. N. and Sadun, A. A. (2002). Optic nerve
degeneration and mitochondrial dysfunction: genetic and acquired optic
neuropathies. Neurochem Int 40, 573-84.
Carter-Dawson, L. D. and LaVail, M. M. (1979). Rods and cones in the mouse
retina. II. Autoradiographic analysis of cell generation using tritiated thymidine. J
Comp Neurol 188, 263-72.
Cayouette, M., Barres, B. A. and Raff, M. (2003). Importance of intrinsic
mechanisms in cell fate decisions in the developing rat retina. Neuron 40, 897904.
Cech, T. R. (1986). The generality of self-splicing RNA: relationship to nuclear
mRNA splicing. Cell 44, 207-10.
Cepko, C. L. (1999). The roles of intrinsic and extrinsic cues and bHLH genes in
the determination of retinal cell fates. Curr Opin Neurobiol 9, 37-46.
Cepko, C. L., Austin, C. P., Yang, X., Alexiades, M. and Ezzeddine, D. (1996).
Cell fate determination in the vertebrate retina. Proc Natl Acad Sci U S A 93,
589-95.
Challa, P. (2004). Glaucoma genetics: advancing new understandings of
glaucoma pathogenesis. Int Ophthalmol Clin 44, 167-85.

303

Chatelain, G., Fossat, N., Brun, G. and Lamonerie, T. (2006). Molecular


dissection reveals decreased activity and not dominant negative effect in human
OTX2 mutants. J Mol Med (Berl) 84, 604-15.
Chen, J. and Smith, L. E. (2007). Retinopathy of prematurity. Angiogenesis 10,
133-40.
Chen, J. H., Wang, D., Huang, C., Zheng, Y., Chen, H., Pang, C. P. and
Zhang, M. (2012). Interactive effects of ATOH7 and RFTN1 in association with
adult-onset primary open angle glaucoma. Invest Ophthalmol Vis Sci.
Chen, P., Johnson, J. E., Zoghbi, H. Y. and Segil, N. (2002). The role of Math1
in inner ear development: Uncoupling the establishment of the sensory
primordium from hair cell fate determination. Development 129, 2495-505.
Chen, S., Wang, Q. L., Nie, Z., Sun, H., Lennon, G., Copeland, N. G., Gilbert,
D. J., Jenkins, N. A. and Zack, D. J. (1997). Crx, a novel Otx-like pairedhomeodomain protein, binds to and transactivates photoreceptor cell-specific
genes. Neuron 19, 1017-30.
Cheng, H., Aleman, T. S., Cideciyan, A. V., Khanna, R., Jacobson, S. G. and
Swaroop, A. (2006). In vivo function of the orphan nuclear receptor NR2E3 in
establishing photoreceptor identity during mammalian retinal development. Hum
Mol Genet 15, 2588-602.
Chien, C. T., Hsiao, C. D., Jan, L. Y. and Jan, Y. N. (1996). Neuronal type
information encoded in the basic-helix-loop-helix domain of proneural genes.
Proc Natl Acad Sci USA 93, 13239-44.
Cho, E. A. and Dressler, G. R. (1998). TCF-4 binds beta-catenin and is
expressed in distinct regions of the embryonic brain and limbs. Mech Dev 77, 918.
Chooniedass-Kothari, S., Emberley, E., Hamedani, M. K., Troup, S., Wang,
X., Czosnek, A., Hube, F., Mutawe, M., Watson, P. H. and Leygue, E. (2004).
The steroid receptor RNA activator is the first functional RNA encoding a protein.
FEBS Lett 566, 43-7.
Chow, R. L. and Lang, R. A. (2001). Early eye development in vertebrates.
Annu Rev Cell Dev Biol 17, 255-96.
Cocquet, J., Chong, A., Zhang, G. and Veitia, R. A. (2006). Reverse
transcriptase template switching and false alternative transcripts. Genomics 88,
127-31.
Cogan, D. G. (1971). Congenital anomalies of the retina. Birth Defects Orig Artic
Ser 7, 41-51.

304

Cohen, J. (1960). A coefficient of agreement for nominal scales. Educ Psychol


Meas 20 3746.
Copt, R. P., Thomas, R. and Mermoud, A. (1999). Corneal thickness in ocular
hypertension, primary open-angle glaucoma, and normal tension glaucoma. Arch
Ophthalmol 117, 14-6.
Corbel, S. Y. and Rossi, F. M. (2002). Latest developments and in vivo use of
the Tet system: ex vivo and in vivo delivery of tetracycline-regulated genes. Curr
Opin Biotechnol 13, 448-52.
Crespo, D., O'Leary, D. D. and Cowan, W. M. (1985). Changes in the numbers
of optic nerve fibers during late prenatal and postnatal development in the albino
rat. Brain Res 351, 129-34.
Cui, S., Otten, C., Rohr, S., Abdelilah-Seyfried, S. and Link, B. A. (2007).
Analysis of aPKClambda and aPKCzeta reveals multiple and redundant functions
during vertebrate retinogenesis. Mol Cell Neurosci 34, 431-44.
Cushman, L. J., Burrows, H. L., Seasholtz, A. F., Lewandoski, M., Muzyczka,
N. and Camper, S. A. (2000). Cre-mediated recombination in the pituitary gland.
Genesis 28, 167-74.
Dakubo, G. D. and Wallace, V. A. (2004). Hedgehogs and retinal ganglion cells:
organizers of the mammalian retina. Neuroreport 15, 479-82.
Dakubo, G. D., Wang, Y. P., Mazerolle, C., Campsall, K., McMahon, A. P. and
Wallace, V. A. (2003). Retinal ganglion cell-derived sonic hedgehog signaling is
required for optic disc and stalk neuroepithelial cell development. Development
130, 2967-80.
Das, G., Choi, Y., Sicinski, P. and Levine, E. M. (2009). Cyclin D1 fine-tunes
the neurogenic output of embryonic retinal progenitor cells. Neural Dev 4, 15.
Das, T., Payer, B., Cayouette, M. and Harris, W. A. (2003). In vivo time-lapse
imaging of cell divisions during neurogenesis in the developing zebrafish retina.
Neuron 37, 597-609.
Dateki, S., Kosaka, K., Hasegawa, K., Tanaka, H., Azuma, N., Yokoya, S.,
Muroya, K., Adachi, M., Tajima, T., Motomura, K. et al. (2010). Heterozygous
orthodenticle homeobox 2 mutations are associated with variable pituitary
phenotype. J Clin Endocrinol Metab 95, 756-64.
Dattani, M. T., Martinez-Barbera, J. P., Thomas, P. Q., Brickman, J. M.,
Gupta, R., Martensson, I. L., Toresson, H., Fox, M., Wales, J. K., Hindmarsh,
P. C. et al. (1998). Mutations in the homeobox gene HESX1/Hesx1 associated
with septo-optic dysplasia in human and mouse. Nat Genet 19, 125-33.

305

De Morsier, G. (1962). Median craioencephalic dysraphias and olfactogenital


dysplasia. World Neurol 3, 485-506.
Del Bene, F., Ettwiller, L., Skowronska-Krawczyk, D., Baier, H., Matter, J. M.,
Birney, E. and Wittbrodt, J. (2007). In vivo validation of a computationally
predicted conserved Ath5 target gene set. PLoS Genet 3, 1661-71.
Derjaguin, B. V. and Churaev, N. V. (1973). Nature of 'anomalous' water.
Nature 244, 430-431.
Desai, A. R. and McConnell, S. K. (2000). Progressive restriction in fate
potential by neural progenitors during cerebral cortical development.
Development 127, 2863-72.
Desgraz, R. and Herrera, P. L. (2009). Pancreatic neurogenin 3-expressing
cells are unipotent islet precursors. Development 136, 3567-74.
Dogan, R. I., Getoor, L., Wilbur, W. J. and Mount, S. M. (2007). SplicePort--an
interactive splice-site analysis tool. Nucleic Acids Res 35, W285-91.
Dokucu, M. E., Zipursky, S. L. and Cagan, R. L. (1996). Atonal, rough and the
resolution of proneural clusters in the developing Drosophila retina. Development
122, 4139-47.
Donovan, S. L. and Dyer, M. A. (2006). Preparation and square wave
electroporation of retinal explant cultures. Nat Protoc 1, 2710-8.
Dorrell, M. I., Aguilar, E. and Friedlander, M. (2002). Retinal vascular
development is mediated by endothelial filopodia, a preexisting astrocytic
template and specific R-cadherin adhesion. Invest Ophthalmol Vis Sci 43, 350010.
Drager, U. C. (1985). Birth dates of retinal ganglion cells giving rise to the
crossed and uncrossed optic projections in the mouse. Proc R Soc Lond B Biol
Sci 224, 57-77.
Dyer, M. A. and Cepko, C. L. (2000). p57(Kip2) regulates progenitor cell
proliferation and amacrine interneuron development in the mouse retina.
Development 127, 3593-605.
Dyer, M. A. and Cepko, C. L. (2001a). p27Kip1 and p57Kip2 regulate
proliferation in distinct retinal progenitor cell populations. J Neurosci 21, 4259-71.
Dyer, M. A. and Cepko, C. L. (2001b). Regulating proliferation during retinal
development. Nat Rev Neurosci 2, 333-42.

306

Dyer, M. A., Livesey, F. J., Cepko, C. L. and Oliver, G. (2003). Prox1 function
controls progenitor cell proliferation and horizontal cell genesis in the mammalian
retina. Nat Genet 34, 53-8.
Dymecki, S. M., Rodriguez, C. I. and Awatramani, R. B. (2002). Switching on
lineage tracers using site-specific recombination. In Embryonic Stem Cells:
Methods and Protocols, vol. 185 (ed. K. Turksen), pp. 309-34: Humana Press.
Echelard, Y., Vassileva, G. and McMahon, A. P. (1994). Cis-acting regulatory
sequences governing Wnt-1 expression in the developing mouse CNS.
Development 120, 2213-24.
Edwards, M. M., McLeod, D. S., Li, R., Grebe, R., Bhutto, I., Mu, X. and Lutty,
G. A. (2011). The deletion of Math5 disrupts retinal blood vessel and glial
development in mice. Exp Eye Res.
Eiraku, M., Takata, N., Ishibashi, H., Kawada, M., Sakakura, E., Okuda, S.,
Sekiguchi, K., Adachi, T. and Sasai, Y. (2011). Self-organizing optic-cup
morphogenesis in three-dimensional culture. Nature 472, 51-6.
Eisenstein, M. (2005). A look back: making mapping easy to digest. Nat
Methods 2, 396.
Ek, U., Fernell, E. and Jacobson, L. (2005). Cognitive and behavioural
characteristics in blind children with bilateral optic nerve hypoplasia. Acta
Paediatr 94, 1421-6.
Elliott, J., Jolicoeur, C., Ramamurthy, V. and Cayouette, M. (2008). Ikaros
confers early temporal competence to mouse retinal progenitor cells. Neuron 60,
26-39.
Elshatory, Y., Deng, M., Xie, X. and Gan, L. (2007a). Expression of the LIMhomeodomain protein Isl1 in the developing and mature mouse retina. J Comp
Neurol 503, 182-97.
Elshatory, Y., Everhart, D., Deng, M., Xie, X., Barlow, R. B. and Gan, L.
(2007b). Islet-1 controls the differentiation of retinal bipolar and cholinergic
amacrine cells. J Neurosci 27, 12707-20.
Enge, M., Bjarnegard, M., Gerhardt, H., Gustafsson, E., Kalen, M., Asker, N.,
Hammes, H. P., Shani, M., Fassler, R. and Betsholtz, C. (2002). Endotheliumspecific platelet-derived growth factor-B ablation mimics diabetic retinopathy.
Embo J 21, 4307-16.
Engelbrecht, J., Knudsen, S. and Brunak, S. (1992). G+C-rich tract in 5' end of
human introns. J Mol Biol 227, 108-13.

307

Erkman, L., McEvilly, R. J., Luo, L., Ryan, A. K., Hooshmand, F., O'Connell,
S. M., Keithley, E. M., Rapaport, D. H., Ryan, A. F. and Rosenfeld, M. G.
(1996). Role of transcription factors Brn-3.1 and Brn-3.2 in auditory and visual
system development. Nature 381, 603-6.
Erkman, L., Yates, P. A., McLaughlin, T., McEvilly, R. J., Whisenhunt, T.,
O'Connell, S. M., Krones, A. I., Kirby, M. A., Rapaport, D. H., Bermingham, J.
R. et al. (2000). A POU domain transcription factor-dependent program regulates
axon pathfinding in the vertebrate visual system. Neuron 28, 779-92.
Erskine, L. and Herrera, E. (2007). The retinal ganglion cell axon's journey:
insights into molecular mechanisms of axon guidance. Dev Biol 308, 1-14.
Esumi, N., Kachi, S., Hackler, L., Jr., Masuda, T., Yang, Z., Campochiaro, P.
A. and Zack, D. J. (2009). BEST1 expression in the retinal pigment epithelium is
modulated by OTX family members. Hum Mol Genet 18, 128-41.
Ezzeddine, Z. D., Yang, X., DeChiara, T., Yancopoulos, G. and Cepko, C. L.
(1997). Postmitotic cells fated to become rod photoreceptors can be respecified
by CNTF treatment of the retina. Development 124, 1055-67.
Fan, B. J., Wang, D. Y., Pasquale, L. R., Haines, J. L. and Wiggs, J. L. (2011).
Genetic variants associated with optic nerve vertical cup-to-disc ratio are risk
factors for primary open angle glaucoma in a US Caucasian population. Invest
Ophthalmol Vis Sci 52, 1788-92.
Fantl, V., Stamp, G., Andrews, A., Rosewell, I. and Dickson, C. (1995). Mice
lacking cyclin D1 are small and show defects in eye and mammary gland
development. Genes Dev 9, 2364-72.
Farah, M., Olson, J., Sucic, H., Hume, R., Tapscott, S. and Turner, D. (2000).
Generation of neurons by transient expression of neural bHLH proteins in
mammalian cells. Development 127, 693-702.
Farah, M. H. and Easter, S. S., Jr. (2005). Cell birth and death in the mouse
retinal ganglion cell layer. J Comp Neurol 489, 120-34.
Fawcett, J. W., O'Leary, D. D. and Cowan, W. M. (1984). Activity and the
control of ganglion cell death in the rat retina. Proc Natl Acad Sci U S A 81, 558993.
Fedorov, A., Saxonov, S., Fedorova, L. and Daizadeh, I. (2001). Comparison
of intron-containing and intron-lacking human genes elucidates putative exonic
splicing enhancers. Nucleic Acids Res 29, 1464-9.
Feinberg, A. P. and Vogelstein, B. (1983). A technique for radiolabeling DNA
restriction endonuclease fragments to high specific activity. Anal Biochem 132, 613.
308

Feng, L., Eisenstat, D. D., Chiba, S., Ishizaki, Y., Gan, L. and Shibasaki, K.
(2011). Brn-3b inhibits generation of amacrine cells by binding to and negatively
regulating DLX1/2 in developing retina. Neuroscience 195, 9-20.
Feng, L., Xie, Z. H., Ding, Q., Xie, X., Libby, R. T. and Gan, L. (2010). MATH5
controls the acquisition of multiple retinal cell fates. Mol Brain 3, 36.
Fisher, R. A. (1925). Statistical Methods for Research Workers: Oliver & Boyd.
Flora, A., Garcia, J. J., Thaller, C. and Zoghbi, H. Y. (2007). The E-protein
Tcf4 interacts with Math1 to regulate differentiation of a specific subset of
neuronal progenitors. Proc Natl Acad Sci U S A 104, 15382-7.
Fox-Walsh, K. L., Dou, Y., Lam, B. J., Hung, S. P., Baldi, P. F. and Hertel, K.
J. (2005). The architecture of pre-mRNAs affects mechanisms of splice-site
pairing. Proc Natl Acad Sci U S A 102, 16176-81.
Frankfort, B. J. and Mardon, G. (2002). R8 development in the Drosophila eye:
a paradigm for neural selection and differentiation. Development 129, 1295-306.
Frohman, M. A. (1993). Rapid amplification of complementary DNA ends for
generation of full-length complementary DNAs: thermal RACE. Methods Enzymol
218, 340-56.
Frohman, M. A., Dush, M. K. and Martin, G. R. (1988). Rapid production of fulllength cDNAs from rare transcripts: amplification using a single gene-specific
oligonucleotide primer. Proc Natl Acad Sci U S A 85, 8998-9002.
Fruttiger, M. (2002). Development of the mouse retinal vasculature:
angiogenesis versus vasculogenesis. Invest Ophthalmol Vis Sci 43, 522-7.
Fruttiger, M. (2007). Development of the retinal vasculature. Angiogenesis 10,
77-88.
Fruttiger, M., Calver, A. R., Kruger, W. H., Mudhar, H. S., Michalovich, D.,
Takakura, N., Nishikawa, S. and Richardson, W. D. (1996). PDGF mediates a
neuron-astrocyte interaction in the developing retina. Neuron 17, 1117-31.
Fruttiger, M., Calver, A. R. and Richardson, W. D. (2000). Platelet-derived
growth factor is constitutively secreted from neuronal cell bodies but not from
axons. Curr Biol 10, 1283-6.
Fu, X., Kiyama, T., Li, R., Russell, M., Klein, W. H. and Mu, X. (2009). Epitopetagging Math5 and Pou4f2: new tools to study retinal ganglion cell development
in the mouse. Dev Dyn 238, 2309-17.

309

Fuhrmann, S., Kirsch, M. and Hofmann, H. D. (1995). Ciliary neurotrophic


factor promotes chick photoreceptor development in vitro. Development 121,
2695-706.
Fujita, S. (1962). Kinetics of cellular proliferation. Exp Cell Res 28, 52-60.
Fujitani, Y., Fujitani, S., Luo, H., Qiu, F., Burlison, J., Long, Q., Kawaguchi,
Y., Edlund, H., MacDonald, R. J., Furukawa, T. et al. (2006). Ptf1a determines
horizontal and amacrine cell fates during mouse retinal development.
Development 133, 4439-50.
Furukawa, A., Koike, C., Lippincott, P., Cepko, C. L. and Furukawa, T.
(2002). The mouse Crx 5'-upstream transgene sequence directs cell-specific and
developmentally regulated expression in retinal photoreceptor cells. J Neurosci
22, 1640-7.
Furukawa, T., Morrow, E. M. and Cepko, C. L. (1997). Crx, a novel otx-like
homeobox gene, shows photoreceptor-specific expression and regulates
photoreceptor differentiation. Cell 91, 531-41.
Furukawa, T., Morrow, E. M., Li, T., Davis, F. C. and Cepko, C. L. (1999).
Retinopathy and attenuated circadian entrainment in Crx-deficient mice. Nat
Genet 23, 466-70.
Furukawa, T., Mukherjee, S., Bao, Z. Z., Morrow, E. M. and Cepko, C. L.
(2000). rax, Hes1, and notch1 promote the formation of Muller glia by postnatal
retinal progenitor cells. Neuron 26, 383-94.
Galli-Resta, L. and Ensini, M. (1996). An intrinsic time limit between genesis
and death of individual neurons in the developing retinal ganglion cell layer. J
Neurosci 16, 2318-24.
Gan, L., Wang, S. W., Huang, Z. and Klein, W. H. (1999). POU domain factor
Brn-3b is essential for retinal ganglion cell differentiation and survival but not for
initial cell fate specification. Dev Biol 210, 469-80.
Gan, L., Xiang, M., Zhou, L., Wagner, D. S., Klein, W. H. and Nathans, J.
(1996). POU domain factor Brn-3b is required for the development of a large set
of retinal ganglion cells. Proc Natl Acad Sci U S A 93, 3920-5.
Gariano, R. F. and Gardner, T. W. (2005). Retinal angiogenesis in development
and disease. Nature 438, 960-6.
Garrick, D., Fiering, S., Martin, D. I. and Whitelaw, E. (1998). Repeat-induced
gene silencing in mammals. Nat Genet 18, 56-9.
Gentles, A. J. and Karlin, S. (1999). Why are human G-protein-coupled
receptors predominantly intronless? Trends Genet 15, 47-9.
310

Georgi, S. A. and Reh, T. A. (2010). Dicer is required for the transition from
early to late progenitor state in the developing mouse retina. J Neurosci 30,
4048-61.
Gerfen, C. R., Holmes, A., Sibley, D., Skolnick, P., Wray, S. and (Eds.).
(2001). Common Stock Solutions, Buffers, and Media. In Current Protocols in
Neuroscience: John Wiley & Sons, Inc.
Ghelli, A., Zanna, C., Porcelli, A. M., Schapira, A. H., Martinuzzi, A., Carelli,
V. and Rugolo, M. (2003). Leber's hereditary optic neuropathy (LHON)
pathogenic mutations induce mitochondrial-dependent apoptotic death in
transmitochondrial cells incubated with galactose medium. J Biol Chem 278,
4145-50.
Ghiasvand, N. M., Fleming, T. P., Helms, C., Avisa, A. and Donis-Keller, H.
(2000). Genetic fine mapping of the gene for nonsyndromic congenital retinal
nonattachment. Am J Med Genet 92, 220-3.
Ghiasvand, N. M., Rudolph, D. D., Mashayekhi, M., Brzezinski, J. A.,
Goldman, D. and Glaser, T. (2011). Deletion of a remote enhancer near ATOH7
disrupts retinal neurogenesis, causing NCRNA disease. Nat Neurosci 14, 57886.
Ghiasvand, N. M., Shirzad, E., Naghavi, M. and Vaez Mahdavi, M. R. (1998).
High incidence of autosomal recessive nonsyndromal congenital retinal
nonattachment (NCRNA) in an Iranian founding population. Am J Med Genet 78,
226-32.
Ghosh, K. K., Bujan, S., Haverkamp, S., Feigenspan, A. and Wassle, H.
(2004). Types of bipolar cells in the mouse retina. J Comp Neurol 469, 70-82.
Glessner, J. T., Wang, K., Cai, G., Korvatska, O., Kim, C. E., Wood, S.,
Zhang, H., Estes, A., Brune, C. W., Bradfield, J. P. et al. (2009). Autism
genome-wide copy number variation reveals ubiquitin and neuronal genes.
Nature 459, 569-73.
Goldberg, M. F. (1997). Persistent fetal vasculature (PFV): an integrated
interpretation of signs and symptoms associated with persistent hyperplastic
primary vitreous (PHPV). LIV Edward Jackson Memorial Lecture. Am J
Ophthalmol 124, 587-626.
Goldowitz, D., Rice, D. S. and Williams, R. W. (1996). Clonal architecture of
the mouse retina. Prog Brain Res 108, 3-15.
Gomes, F. L., Zhang, G., Carbonell, F., Correa, J. A., Harris, W. A., Simons,
B. D. and Cayouette, M. (2011). Reconstruction of rat retinal progenitor cell
lineages in vitro reveals a surprising degree of stochasticity in cell fate decisions.
Development 138, 227-35.
311

Gong, S., Yang, X. W., Li, C. and Heintz, N. (2002). Highly efficient modification
of bacterial artificial chromosomes (BACs) using novel shuttle vectors containing
the R6Kgamma origin of replication. Genome Res 12, 1992-8.
Gooley, J. J., Lu, J., Chou, T. C., Scammell, T. E. and Saper, C. B. (2001).
Melanopsin in cells of origin of the retinohypothalamic tract. Nat Neurosci 4,
1165.
Gotz, M. and Huttner, W. B. (2005). The cell biology of neurogenesis. Nat Rev
Mol Cell Biol 6, 777-88.
Gown, A. M. and Willingham, M. C. (2002). Improved detection of apoptotic
cells in archival paraffin sections: immunohistochemistry using antibodies to
cleaved caspase 3. J Histochem Cytochem 50, 449-54.
Gray, N. K. and Hentze, M. W. (1994). Regulation of protein synthesis by mRNA
structure. Mol Biol Rep 19, 195-200.
Gregory-Evans, C. Y., Williams, M. J., Halford, S. and Gregory-Evans, K.
(2004). Ocular coloboma: a reassessment in the age of molecular neuroscience.
J Med Genet 41, 881-91.
Gutierrez, C., McNally, M. and Canto-Soler, M. V. (2011). Cytoskeleton
proteins previously considered exclusive to ganglion cells are transiently
expressed by all retinal neuronal precursors. BMC Dev Biol 11, 46.
Haddad, N. G. and Eugster, E. A. (2005). Hypopituitarism and
neurodevelopmental abnormalities in relation to central nervous system structural
defects in children with optic nerve hypoplasia. J Pediatr Endocrinol Metab 18,
853-8.
Haddad, R., Font, R. L. and Reeser, F. (1978). Persistent hyperplastic primary
vitreous. A clinicopathologic study of 62 cases and review of the literature. Surv
Ophthalmol 23, 123-34.
Harlow, E. and Lane, D. (1988). Antibodies. A laboratory manual. Cold Spring
Harbor, NY: CSH Press.
Harrington, E. D., Boue, S., Valcarcel, J., Reich, J. G. and Bork, P. (2004).
Estimating rates of alternative splicing in mammals and invertebrates: authors
reply. Nat Genet 36, 916-7.
Harris, W. A. and Hartenstein, V. (1991). Neuronal determination without cell
division in Xenopus embryos. Neuron 6, 499-515.
Hatakeyama, J. and Kageyama, R. (2002). Retrovirus-mediated gene transfer
to retinal explants. Methods 28, 387-95.

312

Hattar, S., Liao, H. W., Takao, M., Berson, D. M. and Yau, K. W. (2002).
Melanopsin-containing retinal ganglion cells: architecture, projections, and
intrinsic photosensitivity. Science 295, 1065-70.
Hawley, R. G. (1994). High-titer retroviral vectors for efficient transduction of
functional genes into murine hematopoietic stem cells. Ann N Y Acad Sci 716,
327-30.
Hayden, S. A., Mills, J. W. and Masland, R. M. (1980). Acetylcholine synthesis
by displaced amacrine cells. Science 210, 435-7.
Heintz, N. (2001). BAC to the future: the use of bac transgenic mice for
neuroscience research. Nat Rev Neurosci 2, 861-70.
Hellman, L. M. and Fried, M. G. (2007). Electrophoretic mobility shift assay
(EMSA) for detecting protein-nucleic acid interactions. Nat Protoc 2, 1849-61.
Helms, A. W., Abney, A. L., Ben-Arie, N., Zoghbi, H. Y. and Johnson, J. E.
(2000). Autoregulation and multiple enhancers control Math1 expression in the
developing nervous system. Development 127, 1185-96.
Henke, W., Herdel, K., Jung, K., Schnorr, D. and Loening, S. A. (1997).
Betaine improves the PCR amplification of GC-rich DNA sequences. Nucleic
Acids Res 25, 3957-8.
Hennig, A. K., Peng, G. H. and Chen, S. (2008). Regulation of photoreceptor
gene expression by Crx-associated transcription factor network. Brain Res 1192,
114-33.
Hertel, K. J. (2008). Combinatorial control of exon recognition. J Biol Chem 283,
1211-5.
Holt, C. E., Bertsch, T. W., Ellis, H. M. and Harris, W. A. (1988). Cellular
determination in the Xenopus retina is independent of lineage and birth date.
Neuron 1, 15-26.
Howell, N. (1998). Leber hereditary optic neuropathy: respiratory chain
dysfunction and degeneration of the optic nerve. Vision Res 38, 1495-504.
Hsiung, F. and Moses, K. (2002). Retinal development in Drosophila: specifying
the first neuron. Hum Mol Genet 11, 1207-14.
Hu, M., Krause, D., Greaves, M., Sharkis, S., Dexter, M., Heyworth, C. and
Enver, T. (1997). Multilineage gene expression precedes commitment in the
hemopoietic system. Genes Dev 11, 774-85.

313

Hufnagel, R. B., Le, T. T., Riesenberg, A. L. and Brown, N. L. (2010). Neurog2


controls the leading edge of neurogenesis in the mammalian retina. Dev Biol
340, 490-503.
Hufnagel, R. B., Riesenberg, A. N., Saul, S. M. and Brown, N. L. (2007).
Conserved regulation of Math5 and Math1 revealed by Math5-GFP transgenes.
Mol Cell Neurosci 36, 435-48.
Hutcheson, D. A., Hanson, M. I., Moore, K. B., Le, T. T., Brown, N. L. and
Vetter, M. L. (2005). bHLH-dependent and -independent modes of Ath5 gene
regulation during retinal development. Development 132, 829-39.
Hutcheson, D. A. and Vetter, M. L. (2001). The bHLH factors Xath5 and
XNeuroD can upregulate the expression of XBrn3d, a POU-homeodomain
transcription factor. Dev Biol 232, 327-38.
Huttner, W. B. and Kosodo, Y. (2005). Symmetric versus asymmetric cell
division during neurogenesis in the developing vertebrate central nervous
system. Curr Opin Cell Biol 17, 648-57.
Irimia, M. and Roy, S. W. (2008). Spliceosomal introns as tools for genomic and
evolutionary analysis. Nucleic Acids Res 36, 1703-12.
Jadhav, A. P., Cho, S. H. and Cepko, C. L. (2006). Notch activity permits retinal
cells to progress through multiple progenitor states and acquire a stem cell
property. Proc Natl Acad Sci U S A 103, 18998-9003.
Jaillon, O., Bouhouche, K., Gout, J. F., Aury, J. M., Noel, B., Saudemont, B.,
Nowacki, M., Serrano, V., Porcel, B. M., Segurens, B. et al. (2008).
Translational control of intron splicing in eukaryotes. Nature 451, 359-62.
James, J., Das, A. V., Bhattacharya, S., Chacko, D. M., Zhao, X. and Ahmad,
I. (2003). In vitro generation of early-born neurons from late retinal progenitors. J
Neurosci 23, 8193-203.
Jameson, B. A. and Wolf, H. (1988). The antigenic index: a novel algorithm for
predicting antigenic determinants. Comput Appl Biosci 4, 181-6.
Jarman, A. P. (2000). Developmental genetics: vertebrates and insects see eye
to eye. Curr Biol 10, R857-9.
Jarman, A. P., Grau, Y., Jan, L. Y. and Jan, Y. N. (1993). atonal is a proneural
gene that directs chordotonal organ formation in the Drosophila peripheral
nervous system. Cell 73, 1307-1321.
Jarman, A. P., Grell, E. H., Ackerman, L., Jan, L. Y. and Jan, Y. N. (1994).
atonal is the proneural gene for Drosophila photoreceptors. Nature 369, 398-400.

314

Jeffares, D. C., Penkett, C. J. and Bahler, J. (2008). Rapidly regulated genes


are intron poor. Trends Genet 24, 375-8.
Jensen, A. M. and Wallace, V. A. (1997). Expression of Sonic hedgehog and its
putative role as a precursor cell mitogen in the developing mouse retina.
Development 124, 363-71.
Jeon, C. J., Strettoi, E. and Masland, R. H. (1998). The major cell populations
of the mouse retina. J Neurosci 18, 8936-46.
Kageyama, R. and Nakanishi, S. (1997). Helix-loop-helix factors in growth and
differentiation of the vertebrate nervous system. Curr Opin Genet Dev 7, 659-65.
Kanadia, R. N. and Cepko, C. L. (2010). Alternative splicing produces high
levels of noncoding isoforms of bHLH transcription factors during development.
Genes Dev 24, 229-34.
Kanekar, S., Perron, M., Dorsky, R., Harris, W. A., Jan, L. Y., Jan, Y. N. and
Vetter, M. L. (1997). Xath5 participates in a network of bHLH genes in the
developing Xenopus retina. Neuron 19, 981-94.
Kay, J. N., Finger-Baier, K. C., Roeser, T., Staub, W. and Baier, H. (2001).
Retinal ganglion cell genesis requires lakritz, a Zebrafish atonal Homolog.
Neuron 30, 725-36.
Kelberman, D. and Dattani, M. T. (2007). Genetics of septo-optic dysplasia.
Pituitary 10, 393-407.
Kenneson, A., Zhang, F., Hagedorn, C. H. and Warren, S. T. (2001). Reduced
FMRP and increased FMR1 transcription is proportionally associated with CGG
repeat number in intermediate-length and premutation carriers. Hum Mol Genet
10, 1449-54.
Key, G., Becker, M. H., Baron, B., Duchrow, M., Schluter, C., Flad, H. D. and
Gerdes, J. (1993). New Ki-67-equivalent murine monoclonal antibodies (MIB 13) generated against bacterially expressed parts of the Ki-67 cDNA containing
three 62 base pair repetitive elements encoding for the Ki-67 epitope. Lab Invest
68, 629-36.
Khaliq, S., Hameed, A., Ismail, M., Anwar, K., Leroy, B., Payne, A. M.,
Bhattacharya, S. S. and Mehdi, S. Q. (2001). Locus for autosomal recessive
nonsyndromic persistent hyperplastic primary vitreous. Invest Ophthalmol Vis Sci
42, 2225-8.
Khan, K., Logan, C. V., McKibbin, M., Sheridan, E., Elcioglu, N. H., Yenice,
O., Parry, D. A., Fernandez-Fuentes, N., Abdelhamed, Z. I., Al-Maskari, A. et
al. (2011). Next generation sequencing identifies mutations in Atonal homolog 7
(ATOH7) in families with global eye developmental defects. Hum Mol Genet.
315

Khor, C. C., Ramdas, W. D., Vithana, E. N., Cornes, B. K., Sim, X., Tay, W. T.,
Saw, S. M., Zheng, Y., Lavanya, R., Wu, R. et al. (2011). Genome-wide
association studies in Asians confirm the involvement of ATOH7 and TGFBR3,
and further identify CARD10 as a novel locus influencing optic disc area. Hum
Mol Genet 20, 1864-72.
Kim, D. S., Ross, S. E., Trimarchi, J. M., Aach, J., Greenberg, M. E. and
Cepko, C. L. (2008). Identification of molecular markers of bipolar cells in the
murine retina. J Comp Neurol 507, 1795-810.
Kim, J., Wu, H. H., Lander, A. D., Lyons, K. M., Matzuk, M. M. and Calof, A. L.
(2005). GDF11 controls the timing of progenitor cell competence in developing
retina. Science 308, 1927-30.
Kitabayashi, M. and Esaka, M. (2003). Improvement of reverse transcription
PCR by RNase H. Biosci Biotechnol Biochem 67, 2474-6.
Kiyama, T., Mao, C. A., Cho, J. H., Fu, X., Pan, P., Mu, X. and Klein, W. H.
(2011). Overlapping spatiotemporal patterns of regulatory gene expression are
required for neuronal progenitors to specify retinal ganglion cell fate. Vision Res
51, 251-9.
Koike, C., Nishida, A., Akimoto, K., Nakaya, M. A., Noda, T., Ohno, S. and
Furukawa, T. (2005). Function of atypical protein kinase C lambda in
differentiating photoreceptors is required for proper lamination of mouse retina. J
Neurosci 25, 10290-8.
Kolpak, A., Zhang, J. and Bao, Z. Z. (2005). Sonic hedgehog has a dual effect
on the growth of retinal ganglion axons depending on its concentration. J
Neurosci 25, 3432-41.
Kriss, J. P. and Revesz, L. (1962). The distribution and fate of
bromodeoxyuridine and bromodeoxycytidine in the mouse and rat. Cancer Res
22, 254-65.
Kumar, S. and Nussinov, R. (2002). Close-range electrostatic interactions in
proteins. Chembiochem 3, 604-17.
Lahav, M., Albert, D. M. and Wyand, S. (1973). Clinical and histopathologic
classification of retinal dysplasia. Am J Ophthalmol 75, 648-67.
Lamba, D. A., Gust, J. and Reh, T. A. (2009). Transplantation of human
embryonic stem cell-derived photoreceptors restores some visual function in Crxdeficient mice. Cell Stem Cell 4, 73-9.
Lamba, D. A., Karl, M. O., Ware, C. B. and Reh, T. A. (2006). Efficient
generation of retinal progenitor cells from human embryonic stem cells. Proc Natl
Acad Sci U S A 103, 12769-74.
316

Lambert, S. R., Hoyt, C. S. and Narahara, M. H. (1987). Optic nerve


hypoplasia. Surv Ophthalmol 32, 1-9.
Lang, R., Lustig, M., Francois, F., Sellinger, M. and Plesken, H. (1994).
Apoptosis during macrophage-dependent ocular tissue remodelling.
Development 120, 3395-403.
Laterra, J., Guerin, C. and Goldstein, G. W. (1990). Astrocytes induce neural
microvascular endothelial cells to form capillary-like structures in vitro. J Cell
Physiol 144, 204-15.
LaVail, M. M., Faktorovich, E. G., Hepler, J. M., Pearson, K. L., Yasumura, D.,
Matthes, M. T. and Steinberg, R. H. (1991). Basic fibroblast growth factor
protects photoreceptors from light-induced degeneration in albino rats. Ann N Y
Acad Sci 638, 341-7.
Le, T. T., Wroblewski, E., Patel, S., Riesenberg, A. N. and Brown, N. L.
(2006). Math5 is required for both early retinal neuron differentiation and cell
cycle progression. Dev Biol 295, 764-78.
Lee, B. L., Bateman, J. B. and Schwartz, S. D. (1996). Posterior segment
neovascularization associated with optic nerve aplasia. Am J Ophthalmol 122,
131-3.
Lee, D. A. and Higginbotham, E. J. (2005). Glaucoma and its treatment: a
review. Am J Health Syst Pharm 62, 691-9.
Lee, E. C., Yu, D., Martinez de Velasco, J., Tessarollo, L., Swing, D. A.,
Court, D. L., Jenkins, N. A. and Copeland, N. G. (2001). A highly efficient
Escherichia coli-based chromosome engineering system adapted for
recombinogenic targeting and subcloning of BAC DNA. Genomics 73, 56-65.
Lehner, B. (2011). Molecular mechanisms of epistasis within and between
genes. Trends Genet 27, 323-31.
Levine, E. M., Roelink, H., Turner, J. and Reh, T. A. (1997). Sonic hedgehog
promotes rod photoreceptor differentiation in mammalian retinal cells in vitro. J
Neurosci 17, 6277-88.
Leygue, E., Murphy, L., Kuttenn, F. and Watson, P. (1996). Triple primer
polymerase chain reaction. A new way to quantify truncated mRNA expression.
Am J Pathol 148, 1097-103.
Li, Q., Lee, J. A. and Black, D. L. (2007). Neuronal regulation of alternative premRNA splicing. Nat Rev Neurosci 8, 819-31.

317

Li, S., Mo, Z., Yang, X., Price, S. M., Shen, M. M. and Xiang, M. (2004). Foxn4
controls the genesis of amacrine and horizontal cells by retinal progenitors.
Neuron 43, 795-807.
Li, Z., Hu, M., Ochocinska, M. J., Joseph, N. M. and Easter, S. S., Jr. (2000).
Modulation of cell proliferation in the embryonic retina of zebrafish (Danio rerio).
Dev Dyn 219, 391-401.
Lin, B., Wang, S. W. and Masland, R. H. (2004). Retinal ganglion cell type, size,
and spacing can be specified independent of homotypic dendritic contacts.
Neuron 43, 475-85.
Little, L. E., Whitmore, P. V. and Wells, T. W., Jr. (1976). Aplasia of the optic
nerve. J Pediatr Ophthalmol 13, 84-8.
Liu, H. and Naismith, J. H. (2008). An efficient one-step site-directed deletion,
insertion, single and multiple-site plasmid mutagenesis protocol. BMC Biotechnol
8, 91.
Liu, W., Khare, S. L., Liang, X., Peters, M. A., Liu, X., Cepko, C. L. and Xiang,
M. (2000). All Brn3 genes can promote retinal ganglion cell differentiation in the
chick. Development 127, 3237-47.
Liu, W., Mo, Z. and Xiang, M. (2001). The Ath5 proneural genes function
upstream of Brn3 POU domain transcription factor genes to promote retinal
ganglion cell development. Proc Natl Acad Sci U S A 98, 1649-54.
Livak, K. J. and Schmittgen, T. D. (2001). Analysis of relative gene expression
data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method.
Methods 25, 402-8.
Livesey, F. J. and Cepko, C. L. (2001). Vertebrate neural cell-fate
determination: lessons from the retina. Nat Rev Neurosci 2, 109-18.
Lobe, C. G., Koop, K. E., Kreppner, W., Lomeli, H., Gertsenstein, M. and
Nagy, A. (1999). Z/AP, a double reporter for cre-mediated recombination. Dev
Biol 208, 281-92.
Longo, A., Guanga, G. P. and Rose, R. B. (2008). Crystal structure of E47NeuroD1/beta2 bHLH domain-DNA complex: heterodimer selectivity and DNA
recognition. Biochemistry 47, 218-29.
Lu, B., Jan, L. and Jan, Y. N. (2000). Control of cell divisions in the nervous
system: symmetry and asymmetry. Annu Rev Neurosci 23, 531-56.
Luo, L. (2007). Fly MARCM and mouse MADM: genetic methods of labeling and
manipulating single neurons. Brain Res Rev 55, 220-7.

318

MacDonald, R. J., Swift, G. H., Przybyla, A. E. and Chirgwin, J. M. (1987).


Isolation of RNA using guanidinium salts. Methods Enzymol 152, 219-27.
Macgregor, S., Hewitt, A. W., Hysi, P. G., Ruddle, J. B., Medland, S. E.,
Henders, A. K., Gordon, S. D., Andrew, T., McEvoy, B., Sanfilippo, P. G. et
al. (2010). Genome-wide association identifies ATOH7 as a major gene
determining human optic disc size. Hum Mol Genet 19, 2716-24.
MacLaren, R. E., Pearson, R. A., MacNeil, A., Douglas, R. H., Salt, T. E.,
Akimoto, M., Swaroop, A., Sowden, J. C. and Ali, R. R. (2006). Retinal repair
by transplantation of photoreceptor precursors. Nature 444, 203-7.
MacNeil, M. A. and Masland, R. H. (1998). Extreme diversity among amacrine
cells: implications for function. Neuron 20, 971-82.
Mader, R. M., Schmidt, W. M., Sedivy, R., Rizovski, B., Braun, J., Kalipciyan,
M., Exner, M., Steger, G. G. and Mueller, M. W. (2001). Reverse transcriptase
template switching during reverse transcriptase-polymerase chain reaction:
artificial generation of deletions in ribonucleotide reductase mRNA. J Lab Clin
Med 137, 422-8.
Maillard, I., Weng, A. P., Carpenter, A. C., Rodriguez, C. G., Sai, H., Xu, L.,
Allman, D., Aster, J. C. and Pear, W. S. (2004). Mastermind critically regulates
Notch-mediated lymphoid cell fate decisions. Blood 104, 1696-702.
Malecki, M. T., Jhala, U. S., Antonellis, A., Fields, L., Doria, A., Orban, T.,
Saad, M., Warram, J. H., Montminy, M. and Krolewski, A. S. (1999). Mutations
in NEUROD1 are associated with the development of type 2 diabetes mellitus.
Nat Genet 23, 323-8.
Mao, C. A., Kiyama, T., Pan, P., Furuta, Y., Hadjantonakis, A. K. and Klein,
W. H. (2008a). Eomesodermin, a target gene of Pou4f2, is required for retinal
ganglion cell and optic nerve development in the mouse. Development 135, 27180.
Mao, C. A., Wang, S. W., Pan, P. and Klein, W. H. (2008b). Rewiring the retinal
ganglion cell gene regulatory network: Neurod1 promotes retinal ganglion cell
fate in the absence of Math5. Development 135, 3379-88.
Mao, X., Fujiwara, Y., Chapdelaine, A., Yang, H. and Orkin, S. H. (2001).
Activation of EGFP expression by Cre-mediated excision in a new ROSA26
reporter mouse strain. Blood 97, 324-6.
Maquat, L. E. and Li, X. (2001). Mammalian heat shock p70 and histone H4
transcripts, which derive from naturally intronless genes, are immune to
nonsense-mediated decay. RNA 7, 445-56.

319

Margulies, E. H., Kardia, S. L. and Innis, J. W. (2001). Identification and


prevention of a GC content bias in SAGE libraries. Nucleic Acids Res 29, E60-0.
Marquardt, T. (2003). Transcriptional control of neuronal diversification in the
retina. Prog Retin Eye Res 22, 567-77.
Marquardt, T. and Gruss, P. (2002). Generating neuronal diversity in the retina:
one for nearly all. Trends Neurosci 25, 32-8.
Martinez-Morales, J. R., Signore, M., Acampora, D., Simeone, A. and
Bovolenta, P. (2001). Otx genes are required for tissue specification in the
developing eye. Development 128, 2019-30.
Martinou, J. C., Dubois-Dauphin, M., Staple, J. K., Rodriguez, I., Frankowski,
H., Missotten, M., Albertini, P., Talabot, D., Catsicas, S., Pietra, C. et al.
(1994). Overexpression of Bcl-2 in transgenic mice protects neurons from
naturally occurring cell death and experimental ischemia. Neuron 13, 1017-1030.
Masai, I. (2000). [Mechanisms underlying induction and progression of a
neurogenic wave in the zebrafish developing retina]. Tanpakushitsu Kakusan
Koso 45, 2782-90.
Masland, R. H. (1988). Amacrine cells. Trends Neurosci 11, 405-10.
Masland, R. H. (2001). Neuronal diversity in the retina. Curr Opin Neurobiol 11,
431-6.
Mastick, G. S. and Andrews, G. L. (2001). Pax6 Regulates the Identity of
Embryonic Diencephalic Neurons. Mol Cell Neurosci 17, 190-207.
Mathews, D. H., Sabina, J., Zuker, M. and Turner, D. H. (1999). Expanded
sequence dependence of thermodynamic parameters improves prediction of
RNA secondary structure. J Mol Biol 288, 911-40.
Matsuo, I., Kuratani, S., Kimura, C., Takeda, N. and Aizawa, S. (1995). Mouse
Otx2 functions in the formation and patterning of rostral head. Genes Dev 9,
2646-58.
Matter-Sadzinski, L., Matter, J. M., Ong, M. T., Hernandez, J. and Ballivet, M.
(2001). Specification of neurotransmitter receptor identity in developing retina:
the chick ATH5 promoter integrates the positive and negative effects of several
bHLH proteins. Development 128, 217-31.
Mayer, W., Smith, A., Fundele, R. and Haaf, T. (2000). Spatial separation of
parental genomes in preimplantation mouse embryos. J Cell Biol 148, 629-34.

320

McCabe, M. J., Alatzoglou, K. S. and Dattani, M. T. (2011). Septo-optic


dysplasia and other midline defects: the role of transcription factors: HESX1 and
beyond. Best Pract Res Clin Endocrinol Metab 25, 115-24.
McConnell, S. K. and Kaznowski, C. E. (1991). Cell cycle dependence of
laminar determination in developing neocortex. Science 254, 282-5.
McCullough, A. J. and Berget, S. M. (2000). An intronic splicing enhancer binds
U1 snRNPs to enhance splicing and select 5' splice sites. Mol Cell Biol 20, 922535.
McLoon, S. C. and Barnes, R. B. (1989). Early differentiation of retinal ganglion
cells: an axonal protein expressed by premigratory and migrating retinal ganglion
cells. J Neurosci 9, 1424-32.
Melchior, W. B., Jr. and Von Hippel, P. H. (1973). Alteration of the relative
stability of dA-dT and dG-dC base pairs in DNA. Proc Natl Acad Sci U S A 70,
298-302.
Melton, D. A., Krieg, P. A., Rebagliati, M. R., Maniatis, T., Zinn, K. and Green,
M. R. (1984). Efficient in vitro synthesis of biologically active RNA and RNA
hybridization probes from plasmids containing a bacteriophage SP6 promoter.
Nucleic Acids Res 12, 7035-56.
Mercer, T. R., Dinger, M. E. and Mattick, J. S. (2009). Long non-coding RNAs:
insights into functions. Nat Rev Genet 10, 155-9.
Miller, M. W. and Nowakowski, R. S. (1988). Use of bromodeoxyuridineimmunohistochemistry to examine the proliferation, migration and time of origin
of cells in the central nervous system. Brain Res 457, 44-52.
Mitrovich, Q. M. and Anderson, P. (2000). Unproductively spliced ribosomal
protein mRNAs are natural targets of mRNA surveillance in C. elegans. Genes
Dev 14, 2173-84.
Moore, K. B., Schneider, M. L. and Vetter, M. L. (2002). Posttranslational
mechanisms control the timing of bHLH function and regulate retinal cell fate.
Neuron 34, 183-95.
Moshiri, A., Gonzalez, E., Tagawa, K., Maeda, H., Wang, M., Frishman, L. J.
and Wang, S. W. (2008). Near complete loss of retinal ganglion cells in the
math5/brn3b double knockout elicits severe reductions of other cell types during
retinal development. Dev Biol 316, 214-27.
Mu, X., Beremand, P. D., Zhao, S., Pershad, R., Sun, H., Scarpa, A., Liang,
S., Thomas, T. L. and Klein, W. H. (2004). Discrete gene sets depend on POU
domain transcription factor Brn3b/Brn-3.2/POU4f2 for their expression in the
mouse embryonic retina. Development 131, 1197-1210.
321

Mu, X., Fu, X., Beremand, P. D., Thomas, T. L. and Klein, W. H. (2008). Gene
regulation logic in retinal ganglion cell development: Isl1 defines a critical branch
distinct from but overlapping with Pou4f2. Proc Natl Acad Sci U S A 105, 6942-7.
Mu, X., Fu, X., Sun, H., Beremand, P. D., Thomas, T. L. and Klein, W. H.
(2005). A gene network downstream of transcription factor Math5 regulates
retinal progenitor cell competence and ganglion cell fate. Dev Biol 280, 467-81.
Mu, X. and Klein, W. H. (2004). A gene regulatory hierarchy for retinal ganglion
cell specification and differentiation. Semin Cell Dev Biol 15, 115-23.
Mukaddes, N. M., Kilincaslan, A., Kucukyazici, G., Sevketoglu, T. and
Tuncer, S. (2007). Autism in visually impaired individuals. Psychiatry Clin
Neurosci 61, 39-44.
Muranishi, Y., Sato, S., Inoue, T., Ueno, S., Koyasu, T., Kondo, M. and
Furukawa, T. (2010). Gene expression analysis of embryonic photoreceptor
precursor cells using BAC-Crx-EGFP transgenic mouse. Biochem Biophys Res
Commun 392, 317-22.
Muranishi, Y., Terada, K., Inoue, T., Katoh, K., Tsujii, T., Sanuki, R.,
Kurokawa, D., Aizawa, S., Tamaki, Y. and Furukawa, T. (2011). An essential
role for RAX homeoprotein and NOTCH-HES signaling in Otx2 expression in
embryonic retinal photoreceptor cell fate determination. J Neurosci 31, 16792807.
Murre, C., McCaw, P. S., Vaessin, H., Caudy, M., Jan, L. Y., Jan, Y. N.,
Cabrera, C. V., Buskin, J. N., Hauschka, S. D. and Lassar, A. B. (1989).
Interactions between heterologous helix-loop-helix proteins generate complexes
that bind specifically to a common DNA sequence. Cell 58, 537-44.
Mytelka, D. S. and Chamberlin, M. J. (1996). Analysis and suppression of DNA
polymerase pauses associated with a trinucleotide consensus. Nucleic Acids Res
24, 2774-81.
Nadal-Nicolas, F. M., Jimenez-Lopez, M., Sobrado-Calvo, P., Nieto-Lopez,
L., Canovas-Martinez, I., Salinas-Navarro, M., Vidal-Sanz, M. and Agudo, M.
(2009). Brn3a as a marker of retinal ganglion cells: qualitative and quantitative
time course studies in naive and optic nerve-injured retinas. Invest Ophthalmol
Vis Sci 50, 3860-8.
Nagy, A. (2000). Cre recombinase: the universal reagent for genome tailoring.
Genesis 26, 99-109.
Nelson, B. R., Hartman, B. H., Georgi, S. A., Lan, M. S. and Reh, T. A. (2007).
Transient inactivation of Notch signaling synchronizes differentiation of neural
progenitor cells. Dev Biol 304, 479-98.

322

Ng, L., Lu, A., Swaroop, A., Sharlin, D. S. and Forrest, D. (2011). Two
transcription factors can direct three photoreceptor outcomes from rod precursor
cells in mouse retinal development. J Neurosci 31, 11118-25.
Nishida, A., Furukawa, A., Koike, C., Tano, Y., Aizawa, S., Matsuo, I. and
Furukawa, T. (2003). Otx2 homeobox gene controls retinal photoreceptor cell
fate and pineal gland development. Nat Neurosci 6, 1255-63.
Norton, J. D. (2000). ID helix-loop-helix proteins in cell growth, differentiation
and tumorigenesis. J Cell Sci 113 ( Pt 22), 3897-905.
Nurse, P. (2000). A long twentieth century of the cell cycle and beyond. Cell 100,
71-8.
O'Leary, D. D., Fawcett, J. W. and Cowan, W. M. (1986). Topographic targeting
errors in the retinocollicular projection and their elimination by selective ganglion
cell death. J Neurosci 6, 3692-705.
Oh, E. C., Khan, N., Novelli, E., Khanna, H., Strettoi, E. and Swaroop, A.
(2007). Transformation of cone precursors to functional rod photoreceptors by
bZIP transcription factor NRL. Proc Natl Acad Sci U S A 104, 1679-84.
Ohnuma, S. and Harris, W. A. (2003). Neurogenesis and the cell cycle. Neuron
40, 199-208.
Ohnuma, S., Hopper, S., Wang, K. C., Philpott, A. and Harris, W. A. (2002).
Co-ordinating retinal histogenesis: early cell cycle exit enhances early cell fate
determination in the Xenopus retina. Development 129, 2435-46.
Ohnuma, S., Philpott, A., Wang, K., Holt, C. E. and Harris, W. A. (1999).
p27Xic1, a Cdk inhibitor, promotes the determination of glial cells in Xenopus
retina. Cell 99, 499-510.
Ohsawa, R. and Kageyama, R. (2008). Regulation of retinal cell fate
specification by multiple transcription factors. Brain Res 1192, 90-8.
Olichon, A., Guillou, E., Delettre, C., Landes, T., Arnaune-Pelloquin, L.,
Emorine, L. J., Mils, V., Daloyau, M., Hamel, C., Amati-Bonneau, P. et al.
(2006). Mitochondrial dynamics and disease, OPA1. Biochim Biophys Acta 1763,
500-9.
Oliver, E. R., Saunders, T. L., Tarle, S. A. and Glaser, T. (2004). Ribosomal
protein L24 defect in belly spot and tail (Bst), a mouse Minute. Development 131,
3907-20.
Oster, S. F., Deiner, M., Birgbauer, E. and Sretavan, D. W. (2004). Ganglion
cell axon pathfinding in the retina and optic nerve. Semin Cell Dev Biol 15, 12536.
323

Osterfield, M., Egelund, R., Young, L. M. and Flanagan, J. G. (2008).


Interaction of amyloid precursor protein with contactins and NgCAM in the
retinotectal system. Development 135, 1189-99.
Pan, L., Deng, M., Xie, X. and Gan, L. (2008). ISL1 and BRN3B co-regulate the
differentiation of murine retinal ganglion cells. Development 135, 1981-90.
Pan, L., Yang, Z., Feng, L. and Gan, L. (2005). Functional equivalence of Brn3
POU-domain transcription factors in mouse retinal neurogenesis. Development
132, 703-12.
Pear, W. (2001). Transient transfection methods for preparation of high-titer
retroviral supernatants. Curr Protoc Mol Biol Chapter 9, Unit9 11.
Pearson, B. J. and Doe, C. Q. (2003). Regulation of neuroblast competence in
Drosophila. Nature 425, 624-8.
Peichl, L. and Gonzalez-Soriano, J. (1993). Unexpected presence of
neurofilaments in axon-bearing horizontal cells of the mammalian retina. J
Neurosci 13, 4091-100.
Peichl, L. and Gonzalez-Soriano, J. (1994). Morphological types of horizontal
cell in rodent retinae: a comparison of rat, mouse, gerbil, and guinea pig. Vis
Neurosci 11, 501-17.
Pequignot, M. O., Provost, A. C., Salle, S., Taupin, P., Sainton, K. M.,
Marchant, D., Martinou, J. C., Ameisen, J. C., Jais, J. P. and Abitbol, M.
(2003). Major role of BAX in apoptosis during retinal development and in
establishment of a functional postnatal retina. Dev Dyn 228, 231-8.
Perron, M., Kanekar, S., Vetter, M. L. and Harris, W. A. (1998). The genetic
sequence of retinal development in the ciliary margin of the Xenopus eye. Dev
Biol 199, 185-200.
Perry, V. H. and Walker, M. (1980). Amacrine cells, displaced amacrine cells
and interplexiform cells in the retina of the rat. Proc R Soc Lond B Biol Sci 208,
415-31.
Pfeiffer, J. K. and Telesnitsky, A. (2001). Effects of limiting homology at the site
of intermolecular recombinogenic template switching during Moloney murine
leukemia virus replication. J Virol 75, 11263-74.
Pinto, D. Pagnamenta, A. T. Klei, L. Anney, R. Merico, D. Regan, R. Conroy,
J. Magalhaes, T. R. Correia, C. Abrahams, B. S. et al. (2010). Functional
impact of global rare copy number variation in autism spectrum disorders. Nature
466, 368-72.

324

Pittman, A. J., Law, M. Y. and Chien, C. B. (2008). Pathfinding in a large


vertebrate axon tract: isotypic interactions guide retinotectal axons at multiple
choice points. Development 135, 2865-71.
Plouhinec, J. L., Sauka-Spengler, T., Germot, A., Le Mentec, C., Cabana, T.,
Harrison, G., Pieau, C., Sire, J. Y., Veron, G. and Mazan, S. (2003). The
mammalian Crx genes are highly divergent representatives of the Otx5 gene
family, a gnathostome orthology class of orthodenticle-related homeogenes
involved in the differentiation of retinal photoreceptors and circadian entrainment.
Mol Biol Evol 20, 513-21.
Poggi, L., Vitorino, M., Masai, I. and Harris, W. A. (2005). Influences on neural
lineage and mode of division in the zebrafish retina in vivo. J Cell Biol 171, 991-9.
Powell, L. M., Zur Lage, P. I., Prentice, D. R., Senthinathan, B. and Jarman,
A. P. (2004). The proneural proteins Atonal and Scute regulate neural target
genes through different E-box binding sites. Mol Cell Biol 24, 9517-26.
Prasov, L., Brown, N. L. and Glaser, T. (2010). A critical analysis of Atoh7
(Math5) mRNA splicing in the developing mouse retina. PLoS One 5, e12315.
Prasov, L. and Glaser, T. (2012). Dynamic expression of ganglion cell markers
in retinal progenitors during the terminal cell cycle. J Cell Biol (under review).
Prasov, L., Masud, T., Khaliq, S., Mehdi, S. Q., Abid, A., Oliver, E. R., Silva,
E. D., Lewanda, A., Brodsky, M. C., Borchert, M. et al. (2012). ATOH7
mutations cause autosomal recessive persistent hyperplasia of the primary
vitreous. Am J Hum Genet.
Project, N. E. S. Exome Variant Server vol. 2011 (ed. Seattle, WA.
Provencio, I., Jiang, G., De Grip, W. J., Hayes, W. P. and Rollag, M. D.
(1998). Melanopsin: An opsin in melanophores, brain, and eye. Proc Natl Acad
Sci U S A 95, 340-5.
Provis, J. M. (2001). Development of the primate retinal vasculature. Prog Retin
Eye Res 20, 799-821.
Pruett, R. C. (1975). The pleomorphism and complications of posterior
hyperplastic primary vitreous. Am J Ophthalmol 80, 625-9.
Qiu, F., Jiang, H. and Xiang, M. (2008). A comprehensive negative regulatory
program controlled by Brn3b to ensure ganglion cell specification from
multipotential retinal precursors. J Neurosci 28, 3392-403.
Quan, X. J., Denayer, T., Yan, J., Jafar-Nejad, H., Philippi, A., Lichtarge, O.,
Vleminckx, K. and Hassan, B. A. (2004). Evolution of neural precursor

325

selection: functional divergence of proneural proteins. Development 131, 167989.


Quigley, H. A. (1996). Number of people with glaucoma worldwide. Br J
Ophthalmol 80, 389-93.
Rachel, R. A., Dolen, G., Hayes, N. L., Lu, A., Erskine, L., Nowakowski, R. S.
and Mason, C. A. (2002). Spatiotemporal features of early neuronogenesis differ
in wild-type and albino mouse retina. J Neurosci 22, 4249-63.
Ragge, N. K., Brown, A. G., Poloschek, C. M., Lorenz, B., Henderson, R. A.,
Clarke, M. P., Russell-Eggitt, I., Fielder, A., Gerrelli, D., Martinez-Barbera, J.
P. et al. (2005). Heterozygous mutations of OTX2 cause severe ocular
malformations. Am J Hum Genet 76, 1008-22.
Ramdas, W. D., van Koolwijk, L. M., Ikram, M. K., Jansonius, N. M., de Jong,
P. T., Bergen, A. A., Isaacs, A., Amin, N., Aulchenko, Y. S., Wolfs, R. C. et al.
(2010). A genome-wide association study of optic disc parameters. PLoS Genet
6, e1000978.
Ramdas, W. D., van Koolwijk, L. M., Lemij, H. G., Pasutto, F., Cree, A. J.,
Thorleifsson, G., Janssen, S. F., Jacoline, T. B., Amin, N., Rivadeneira, F. et
al. (2011). Common genetic variants associated with open-angle glaucoma. Hum
Mol Genet 20, 2464-71.
Rapaport, D. H., Patheal, S. L. and Harris, W. A. (2001). Cellular competence
plays a role in photoreceptor differentiation in the developing Xenopus retina. J
Neurobiol 49, 129-41.
Rapaport, D. H., Wong, L. L., Wood, E. D., Yasumura, D. and LaVail, M. M.
(2004). Timing and topography of cell genesis in the rat retina. J Comp Neurol
474, 304-24.
Raper, J. and Mason, C. (2010). Cellular strategies of axonal pathfinding. Cold
Spring Harb Perspect Biol 2, a001933.
Rees, W. A., Yager, T. D., Korte, J. and von Hippel, P. H. (1993). Betaine can
eliminate the base pair composition dependence of DNA melting. Biochemistry
32, 137-44.
Reese, A. B. (1955). Persistent hyperplastic primary vitreous. Am J Ophthalmol
40, 317-31.
Reese, B. E. and Colello, R. J. (1992). Neurogenesis in the retinal ganglion cell
layer of the rat. Neuroscience 46, 419-29.
Reese, B. E. and Galli-Resta, L. (2002). The role of tangential dispersion in
retinal mosaic formation. Prog Retin Eye Res 21, 153-68.
326

Reese, B. E., Necessary, B. D., Tam, P. P., Faulkner-Jones, B. and Tan, S. S.


(1999). Clonal expansion and cell dispersion in the developing mouse retina. Eur
J Neurosci 11, 2965-78.
Reh, T. A. (1992). Cellular interactions determine neuronal phenotypes in rodent
retinal cultures. J Neurobiol 23, 1067-83.
Reh, T. A. and Cagan, R. L. (1994). Intrinsic and extrinsic signals in the
developing vertebrate and fly eyes: viewing vertebrate and invertebrate eyes in
the same light. Perspect Dev Neurobiol 2, 183-90.
Reh, T. A. and Kljavin, I. J. (1989). Age of differentiation determines rat retinal
germinal cell phenotype: induction of differentiation by dissociation. J Neurosci 9,
4179-89.
Reh, T. A., Lamba, D. and Gust, J. (2010). Directing human embryonic stem
cells to a retinal fate. Methods Mol Biol 636, 139-53.
Repka, A. M. and Adler, R. (1992). Accurate determination of the time of cell
birth using a sequential labeling technique with [3H]-thymidine and
bromodeoxyuridine ("window labeling"). J Histochem Cytochem 40, 947-53.
Rhodes, K. J. and Trimmer, J. S. (2006). Antibodies as valuable neuroscience
research tools versus reagents of mass distraction. J Neurosci 26, 8017-20.
Riesenberg, A. N., Liu, Z., Kopan, R. and Brown, N. L. (2009). Rbpj cell
autonomous regulation of retinal ganglion cell and cone photoreceptor fates in
the mouse retina. J Neurosci 29, 12865-77.
Robitaille, J. M., Wallace, K., Zheng, B., Beis, M. J., Samuels, M., HoskinMott, A. and Guernsey, D. L. (2009). Phenotypic overlap of familial exudative
vitreoretinopathy (FEVR) with persistent fetal vasculature (PFV) caused by FZD4
mutations in two distinct pedigrees. Ophthalmic Genet 30, 23-30.
Rockhill, R. L., Daly, F. J., MacNeil, M. A., Brown, S. P. and Masland, R. H.
(2002). The diversity of ganglion cells in a mammalian retina. J Neurosci 22,
3831-43.
Rodieck, R. W. (1998). The First Steps in Seeing. Sunderland, MA: Sinauer.
Roe, T., Reynolds, T. C., Yu, G. and Brown, P. O. (1993). Integration of murine
leukemia virus DNA depends on mitosis. Embo J 12, 2099-108.
Rompani, S. B. and Cepko, C. L. (2008). Retinal progenitor cells can produce
restricted subsets of horizontal cells. Proc Natl Acad Sci U S A 105, 192-7.
Roohi, J., Montagna, C., Tegay, D. H., Palmer, L. E., DeVincent, C.,
Pomeroy, J. C., Christian, S. L., Nowak, N. and Hatchwell, E. (2009).
327

Disruption of contactin 4 in three subjects with autism spectrum disorder. J Med


Genet 46, 176-82.
Ross, S. E., Greenberg, M. E. and Stiles, C. D. (2003). Basic helix-loop-helix
factors in cortical development. Neuron 39, 13-25.
Rowan, S. and Cepko, C. L. (2004). Genetic analysis of the homeodomain
transcription factor Chx10 in the retina using a novel multifunctional BAC
transgenic mouse reporter. Dev Biol 271, 388-402.
Roy, S. W. and Irimia, M. (2008). When good transcripts go bad: artifactual RTPCR 'splicing' and genome analysis. Bioessays 30, 601-5.
Rupp, R. A., Snider, L. and Weintraub, H. (1994). Xenopus embryos regulate
the nuclear localization of XMyoD. Genes Dev 8, 1311-23.
Ruvkun, G. (1997). Patterning the Nervous System.
Sadun, A. A. and Carelli, V. (2003). Mitochondrial function and dysfunction
within the optic nerve. Arch Ophthalmol 121, 1342-3.
Sakharkar, M. K., Chow, V. T., Chaturvedi, I., Mathura, V. S., Shapshak, P.
and Kangueane, P. (2004). A report on single exon genes (SEG) in eukaryotes.
Front Biosci 9, 3262-7.
Samson, M., Emerson, M. M. and Cepko, C. L. (2009). Robust marking of
photoreceptor cells and pinealocytes with several reporters under control of the
Crx gene. Dev Dyn 238, 3218-25.
Sanchez-Camacho, C. and Bovolenta, P. (2008). Autonomous and nonautonomous Shh signalling mediate the in vivo growth and guidance of mouse
retinal ganglion cell axons. Development 135, 3531-41.
Saper, C. B. and Sawchenko, P. E. (2003). Magic peptides, magic antibodies:
guidelines for appropriate controls for immunohistochemistry. J Comp Neurol
465, 161-3.
Sato, S., Inoue, T., Terada, K., Matsuo, I., Aizawa, S., Tano, Y., Fujikado, T.
and Furukawa, T. (2007). Dkk3-Cre BAC transgenic mouse line: a tool for highly
efficient gene deletion in retinal progenitor cells. Genesis 45, 502-7.
Saul, S. M., Brzezinski, J. A., Altschuler, R. A., Shore, S. E., Rudolph, D. D.,
Kabara, L. L., Halsey, K. E., Hufnagel, R. B., Zhou, J., Dolan, D. F. et al.
(2008). Math5 expression and function in the central auditory system. Mol Cell
Neurosci 37, 153-69.

328

Scheetz, A. J., Williams, R. W. and Dubin, M. W. (1995). Severity of ganglion


cell death during early postnatal development is modulated by both neuronal
activity and binocular competition. Vis Neurosci 12, 605-10.
Schiller, P. H. and Malpeli, J. G. (1977). Properties and tectal projections of
monkey retinal ganglion cells. J Neurophysiol 40, 428-45.
Schilter, K. F., Schneider, A., Bardakjian, T., Soucy, J. F., Tyler, R. C., Reis,
L. M. and Semina, E. V. (2011). OTX2 microphthalmia syndrome: four novel
mutations and delineation of a phenotype. Clin Genet 79, 158-68.
Schneider-Poetsch, T., Ju, J., Eyler, D. E., Dang, Y., Bhat, S., Merrick, W. C.,
Green, R., Shen, B. and Liu, J. O. (2010). Inhibition of eukaryotic translation
elongation by cycloheximide and lactimidomycin. Nat Chem Biol 6, 209-217.
Schneider, M. L., Turner, D. L. and Vetter, M. L. (2001). Notch signaling can
inhibit Xath5 function in the neural plate and developing retina. Mol Cell Neurosci
18, 458-72.
Schonbrunner, N. J., Fiss, E. H., Budker, O., Stoffel, S., Sigua, C. L.,
Gelfand, D. H. and Myers, T. W. (2006). Chimeric thermostable DNA
polymerases with reverse transcriptase and attenuated 3'-5' exonuclease activity.
Biochemistry 45, 12786-95.
Schwede, T., Kopp, J., Guex, N. and Peitsch, M. C. (2003). SWISS-MODEL:
An automated protein homology-modeling server. Nucleic Acids Res 31, 3381-5.
Scott, I. U., Warman, R. and Altman, N. (1997). Bilateral aplasia of the optic
nerves, chiasm, and tracts in an otherwise healthy infant. Am J Ophthalmol 124,
409-10.
Shastry, B. S. (2009). Persistent hyperplastic primary vitreous: congenital
malformation of the eye. Clin Experiment Ophthalmol 37, 884-90.
Shen, Q., Wang, Y., Dimos, J. T., Fasano, C. A., Phoenix, T. N., Lemischka, I.
R., Ivanova, N. B., Stifani, S., Morrisey, E. E. and Temple, S. (2006). The
timing of cortical neurogenesis is encoded within lineages of individual progenitor
cells. Nat Neurosci 9, 743-51.
Shen, Y. C. and Raymond, P. A. (2004). Zebrafish cone-rod (crx) homeobox
gene promotes retinogenesis. Dev Biol 269, 237-51.
Shimoda, Y. and Watanabe, K. (2009). Contactins: emerging key roles in the
development and function of the nervous system. Cell Adh Migr 3, 64-70.
Shkumatava, A., Fischer, S., Muller, F., Strahle, U. and Neumann, C. J.
(2004). Sonic hedgehog, secreted by amacrine cells, acts as a short-range signal

329

to direct differentiation and lamination in the zebrafish retina. Development 131,


3849-58.
Sidman, R. L. (1961). Histogenesis of mouse retina studied with thymidine-H3.
In Structure of the Eye, (ed. G. K. Smelser), pp. 487-506. New York: Academic
Press.
Silva, A. O., Ercole, C. E. and McLoon, S. C. (2003). Regulation of ganglion
cell production by Notch signaling during retinal development. J Neurobiol 54,
511-24.
Simeone, A., Acampora, D., Mallamaci, A., Stornaiuolo, A., D'Apice, M. R.,
Nigro, V. and Boncinelli, E. (1993). A vertebrate gene related to orthodenticle
contains a homeodomain of the bicoid class and demarcates anterior
neuroectoderm in the gastrulating mouse embryo. Embo J 12, 2735-47.
Simpson, J. I. (1984). The accessory optic system. Annu Rev Neurosci 7, 13-41.
Sinitsina, V. F. (1971). [DNA synthesis and cell population kinetics in embryonal
histogenesis of the retina in mice]. Arkh Anat Gistol Embriol 61, 58-67.
Sloan, S. R., Shen, C. P., McCarrick-Walmsley, R. and Kadesch, T. (1996).
Phosphorylation of E47 as a potential determinant of B-cell-specific activity. Mol
Cell Biol 16, 6900-8.
Snow, R. L. and Robson, J. A. (1994). Ganglion cell neurogenesis, migration
and early differentiation in the chick retina. Neuroscience 58, 399-409.
Spemann, H. (1901). Ueber korrelationen in der entwicklung des auges. Verh.
Anat. Ges. 15, 6179.
Stacey, D. W. (2003). Cyclin D1 serves as a cell cycle regulatory switch in
actively proliferating cells. Curr Opin Cell Biol 15, 158-63.
Stolting, K. N., Gort, G., Wust, C. and Wilson, A. B. (2009). Eukaryotic
transcriptomics in silico: optimizing cDNA-AFLP efficiency. BMC Genomics 10,
565.
Strom, R. C. and Williams, R. W. (1998). Cell production and cell death in the
generation of variation in neuron number. J Neurosci 18, 9948-53.
Sun, Y., Kanekar, S. L., Vetter, M. L., Gorski, S., Jan, Y. N., Glaser, T. and
Brown, N. L. (2003). Conserved and divergent functions of Drosophila atonal,
amphibian, and mammalian Ath5 genes. Evol Dev 5, 532-41.
Swaroop, A., Kim, D. and Forrest, D. (2010). Transcriptional regulation of
photoreceptor development and homeostasis in the mammalian retina. Nat Rev
Neurosci 11, 563-76.
330

Swift, S., Lorens, J., Achacoso, P. and Nolan, G. P. (2001). Rapid production
of retroviruses for efficient gene delivery to mammalian cells using 293T cellbased systems. Curr Protoc Immunol Chapter 10, Unit 10 17C.
Tabaska, J. E. and Zhang, M. Q. (1999). Detection of polyadenylation signals in
human DNA sequences. Gene 231, 77-86.
Takatsuka, K., Hatakeyama, J., Bessho, Y. and Kageyama, R. (2004). Roles
of the bHLH gene Hes1 in retinal morphogenesis. Brain Res 1004, 148-55.
Taylor, D. (2007). Developmental abnormalities of the optic nerve and chiasm.
Eye (Lond) 21, 1271-84.
Telesnitsky, A. and Goff, S. (1997). Reverse transcription and the generation of
retroviral DNA. In Retroviruses, (ed. J. M. Coffin S. H. Hughes and H. E.
Varmus), pp. 121-160. Cold Spring Harbor, NY: CSH Press.
Thomason, L. C., Costantino, N., Shaw, D. V. and Court, D. L. (2007).
Multicopy plasmid modification with phage lambda Red recombineering. Plasmid
58, 148-58.
Trimarchi, J. M., Stadler, M. B. and Cepko, C. L. (2008). Individual retinal
progenitor cells display extensive heterogeneity of gene expression. PLoS One
3, e1588.
Triplett, J. W., Pfeiffenberger, C., Yamada, J., Stafford, B. K., Sweeney, N.
T., Litke, A. M., Sher, A., Koulakov, A. A. and Feldheim, D. A. (2011).
Competition is a driving force in topographic mapping. Proc Natl Acad Sci U S A.
Turner, D. L. and Cepko, C. L. (1987). A common progenitor for neurons and
glia persists in rat retina late in development. Nature 328, 131-6.
Turner, D. L., Snyder, E. Y. and Cepko, C. L. (1990). Lineage-independent
determination of cell type in the embryonic mouse retina. Neuron 4, 833-45.
Turner, D. L. and Weintraub, H. (1994). Expression of achaete-scute homolog 3
in Xenopus embryos converts ectodermal cells to a neural fate. Genes Dev 8,
1434-47.
Van Gelder, R. N., Wee, R., Lee, J. A. and Tu, D. C. (2003). Reduced pupillary
light responses in mice lacking cryptochromes. Science 299, 222.
Van Parijs, L., Refaeli, Y., Lord, J. D., Nelson, B. H., Abbas, A. K. and
Baltimore, D. (1999). Uncoupling IL-2 signals that regulate T cell proliferation,
survival, and Fas-mediated activation-induced cell death. Immunity 11, 281-8.
Vecino, E., Hernandez, M. and Garcia, M. (2004). Cell death in the developing
vertebrate retina. Int J Dev Biol 48, 965-74.
331

Voinescu, P. E., Kay, J. N. and Sanes, J. R. (2009). Birthdays of retinal


amacrine cell subtypes are systematically related to their molecular identity and
soma position. J Comp Neurol 517, 737-50.
von Bohlen und Halbach, O. (1999). The isolated mammalian brain: an in vivo
preparation suitable for pathway tracing. Eur J Neurosci 11, 1096-100.
Voyvodic, J. T., Burne, J. F. and Raff, M. C. (1995). Quantification of normal
cell death in the rat retina: implications for clone composition in cell lineage
analysis. Eur J Neurosci 7, 2469-78.
Wadman, I. A., Osada, H., Grutz, G. G., Agulnick, A. D., Westphal, H.,
Forster, A. and Rabbitts, T. H. (1997). The LIM-only protein Lmo2 is a bridging
molecule assembling an erythroid, DNA-binding complex which includes the
TAL1, E47, GATA-1 and Ldb1/NLI proteins. Embo J 16, 3145-57.
Wahle, E. (1995). Poly(A) tail length control is caused by termination of
processive synthesis. J Biol Chem 270, 2800-8.
Waid, D. K. and McLoon, S. C. (1995). Immediate differentiation of ganglion
cells following mitosis in the developing retina. Neuron 14, 117-24.
Waid, D. K. and McLoon, S. C. (1998). Ganglion cells influence the fate of
dividing retinal cells in culture. Development 125, 1059-66.
Wallace, V. A. and Raff, M. C. (1999). A role for Sonic hedgehog in axon-toastrocyte signalling in the rodent optic nerve. Development 126, 2901-9.
Wang, S. W., Kim, B. S., Ding, K., Wang, H., Sun, D., Johnson, R. L., Klein,
W. H. and Gan, L. (2001). Requirement for math5 in the development of retinal
ganglion cells. Genes Dev 15, 24-9.
Wang, S. W., Mu, X., Bowers, W. J., Kim, D. S., Plas, D. J., Crair, M. C.,
Federoff, H. J., Gan, L. and Klein, W. H. (2002a). Brn3b/Brn3c double knockout
mice reveal an unsuspected role for Brn3c in retinal ganglion cell axon
outgrowth. Development 129, 467-77.
Wang, S. W., Mu, X., Bowers, W. J. and Klein, W. H. (2002b). Retinal ganglion
cell differentiation in cultured mouse retinal explants. Methods 28, 448-56.
Wang, Y., Dakubo, G. D., Thurig, S., Mazerolle, C. J. and Wallace, V. A.
(2005). Retinal ganglion cell-derived sonic hedgehog locally controls proliferation
and the timing of RGC development in the embryonic mouse retina.
Development 132, 5103-13.
Wang, Y., Smallwood, P. M., Cowan, M., Blesh, D., Lawler, A. and Nathans,
J. (1999). Mutually exclusive expression of human red and green visual pigment-

332

reporter transgenes occurs at high frequency in murine cone photoreceptors.


Proc Natl Acad Sci U S A 96, 5251-6.
Wang, Z. and Burge, C. B. (2008). Splicing regulation: from a parts list of
regulatory elements to an integrated splicing code. RNA 14, 802-13.
Warden, S. M., Andreoli, C. M. and Mukai, S. (2007). The Wnt signaling
pathway in familial exudative vitreoretinopathy and Norrie disease. Semin
Ophthalmol 22, 211-7.
Warming, S., Costantino, N., Court, D. L., Jenkins, N. A. and Copeland, N. G.
(2005). Simple and highly efficient BAC recombineering using galK selection.
Nucleic Acids Res 33, e36.
Watanabe, M., Rutishauser, U. and Silver, J. (1991). Formation of the retinal
ganglion cell and optic fiber layers. J Neurobiol 22, 85-96.
Watanabe, T. and Raff, M. C. (1988). Retinal astrocytes are immigrants from the
optic nerve. Nature 332, 834-7.
Watanabe, T. and Raff, M. C. (1990). Rod photoreceptor development in vitro:
intrinsic properties of proliferating neuroepithelial cells change as development
proceeds in the rat retina. Neuron 4, 461-7.
Wee, R., Castrucci, A. M., Provencio, I., Gan, L. and Van Gelder, R. N.
(2002). Loss of photic entrainment and altered free-running circadian rhythms in
math5-/- mice. J Neurosci 22, 10427-33.
Weissensteiner, T. and Lanchbury, J. S. (1996). Strategy for controlling
preferential amplification and avoiding false negatives in PCR typing.
Biotechniques 21, 1102-8.
Wendt, H., Thomas, R. M. and Ellenberger, T. (1998). DNA-mediated folding
and assembly of MyoD-E47 heterodimers. J Biol Chem 273, 5735-43.
Wessells, N. K. (1977). Tissue Interactions and Development. Menlo Park, CA:
W.A. Benjamin.
Wetts, R. and Fraser, S. E. (1988). Multipotent precursors can give rise to all
major cell types of the frog retina. Science 239, 1142-5.
Willbold, E., Rothermel, A., Tomlinson, S. and Layer, P. G. (2000). Muller glia
cells reorganize reaggregating chicken retinal cells into correctly laminated in
vitro retinae. Glia 29, 45-57.
Wodarz, A., Ramrath, A., Grimm, A. and Knust, E. (2000). Drosophila atypical
protein kinase C associates with Bazooka and controls polarity of epithelia and
neuroblasts. J Cell Biol 150, 1361-74.
333

Wong, L. L. and Rapaport, D. H. (2009). Defining retinal progenitor cell


competence in Xenopus laevis by clonal analysis. Development 136, 1707-15.
Wu, F., Sapkota, D., Li, R. and Mu, X. (2012). Onecut 1 and onecut 2 are
potential regulators of mouse retinal development. J Comp Neurol 520, 952-69.
Wu, W., Blumberg, B. M., Fay, P. J. and Bambara, R. A. (1995). Strand
transfer mediated by human immunodeficiency virus reverse transcriptase in vitro
is promoted by pausing and results in misincorporation. J Biol Chem 270, 32532.
Wyatt, A., Bakrania, P., Bunyan, D. J., Osborne, R. J., Crolla, J. A., Salt, A.,
Ayuso, C., Newbury-Ecob, R., Abou-Rayyah, Y., Collin, J. R. et al. (2008).
Novel heterozygous OTX2 mutations and whole gene deletions in anophthalmia,
microphthalmia and coloboma. Hum Mutat 29, E278-83.
Xiang, M. (1998). Requirement for Brn-3b in early differentiation of postmitotic
retinal ganglion cell precursors. Dev Biol 197, 155-69.
Xiang, M., Zhou, L., Macke, J. P., Yoshioka, T., Hendry, S. H., Eddy, R. L.,
Shows, T. B. and Nathans, J. (1995). The Brn-3 family of POU-domain factors:
primary structure, binding specificity, and expression in subsets of retinal
ganglion cells and somatosensory neurons. J Neurosci 15, 4762-85.
Xiang, M., Zhou, L., Peng, Y. W., Eddy, R. L., Shows, T. B. and Nathans, J.
(1993). Brn-3b: a POU domain gene expressed in a subset of retinal ganglion
cells. Neuron 11, 689-701.
Yamada, E. (1969). Some structural features of the fovea centralis in the human
retina. Arch Ophthalmol 82, 151-9.
Yang, K., Hitomi, M. and Stacey, D. W. (2006). Variations in cyclin D1 levels
through the cell cycle determine the proliferative fate of a cell. Cell Div 1, 32.
Yang, X. J. (2004). Roles of cell-extrinsic growth factors in vertebrate eye pattern
formation and retinogenesis. Semin Cell Dev Biol 15, 91-103.
Yang, Z., Ding, K., Pan, L., Deng, M. and Gan, L. (2003). Math5 determines the
competence state of retinal ganglion cell progenitors. Dev Biol 264, 240-54.
Yao, J., Sun, X., Wang, Y., Xu, G. and Qian, J. (2007). Math5 promotes retinal
ganglion cell expression patterns in retinal progenitor cells. Mol Vis 13, 1066-72.
Yaron, O., Farhy, C., Marquardt, T., Applebury, M. and Ashery-Padan, R.
(2006). Notch1 functions to suppress cone-photoreceptor fate specification in the
developing mouse retina. Development 133, 1367-78.

334

Yau, K. W. and Baylor, D. A. (1989). Cyclic GMP-activated conductance of


retinal photoreceptor cells. Annu Rev Neurosci 12, 289-327.
Yoshihara, Y., Kawasaki, M., Tamada, A., Nagata, S., Kagamiyama, H. and
Mori, K. (1995). Overlapping and differential expression of BIG-2, BIG-1, TAG-1,
and F3: four members of an axon-associated cell adhesion molecule subgroup of
the immunoglobulin superfamily. J Neurobiol 28, 51-69.
Young, R. W. (1984). Cell death during differentiation of the retina in the mouse.
J Comp Neurol 229, 362-73.
Young, R. W. (1985a). Cell differentiation in the retina of the mouse. Anat Rec
212, 199-205.
Young, R. W. (1985b). Cell proliferation during postnatal development of the
retina in the mouse. Brain Res 353, 229-39.
Young, T. L. and Cepko, C. L. (2004). A role for ligand-gated ion channels in
rod photoreceptor development. Neuron 41, 867-79.
Yu, C., Mazerolle, C. J., Thurig, S., Wang, Y., Pacal, M., Bremner, R. and
Wallace, V. A. (2006). Direct and indirect effects of hedgehog pathway activation
in the mammalian retina. Mol Cell Neurosci 32, 274-82.
Yuodelis, C. and Hendrickson, A. (1986). A qualitative and quantitative
analysis of the human fovea during development. Vision Res 26, 847-55.
Zambrowicz, B. P., Imamoto, A., Fiering, S., Herzenberg, L. A., Kerr, W. G.
and Soriano, P. (1997). Disruption of overlapping transcripts in the ROSA beta
geo 26 gene trap strain leads to widespread expression of beta-galactosidase in
mouse embryos and hematopoietic cells. Proc Natl Acad Sci U S A 94, 3789-94.
Zaphiropoulos, P. G. (2002). Template switching generated during reverse
transcription? FEBS Lett 527, 326.
Zhang, H., Deo, M., Thompson, R. C., Uhler, M. D. and Turner, D. L. (2012).
Negative regulation of Yap during neuronal differentiation. Dev Biol 361, 103-15.
Zhang, J., Gray, J., Wu, L., Leone, G., Rowan, S., Cepko, C. L., Zhu, X.,
Craft, C. M. and Dyer, M. A. (2004). Rb regulates proliferation and rod
photoreceptor development in the mouse retina. Nat Genet 36, 351-60.
Zhang, X. M. and Yang, X. J. (2001). Regulation of retinal ganglion cell
production by Sonic hedgehog. Development 128, 943-57.
Zhao, C. and Emmons, S. W. (1995). A transcription factor controlling
development of peripheral sense organs in C. elegans. Nature 373, 74-8.

335

Zhou, P. (2004). Determining protein half-lives. Methods Mol Biol 284, 67-77.
Zhu, M., Provis, J. M. and Penfold, P. L. (1999). The human hyaloid system:
cellular phenotypes and inter-relationships. Exp Eye Res 68, 553-63.
Zhu, X. and Craft, C. M. (2000). Modulation of CRX transactivation activity by
phosducin isoforms. Mol Cell Biol 20, 5216-26.
Zong, H., Espinosa, J. S., Su, H. H., Muzumdar, M. D. and Luo, L. (2005).
Mosaic analysis with double markers in mice. Cell 121, 479-92.
Zuker, M. (2003). Mfold web server for nucleic acid folding and hybridization
prediction. Nucleic Acids Res 31, 3406-15.

336

Вам также может понравиться